Addition to the Carbon-Carbon Triple Bond. The Reaction of Acetic Acid with 3-Hexyne Citation Lemaire, Henry (1950) Addition to the Carbon-Carbon Triple Bond. The Reaction of Acetic Acid with 3-Hexyne. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/5g6h-b137. https://resolver.caltech.edu/CaltechTHESIS:10312023-185511760 Abstract The mechanism of the mercury-catalyzed addition to the carbon-carbon triple bond has been investigated. Acetic acid was used as the solvent and one reactant. The other reactant was 3-hexyne. The course of the reaction was followed by bromine titration and by a special indicator method. The addition was found to take place in two discrete steps, forming 3-acetoxyhexene-3 and 3-hexanone. The rate of the initial reaction was determined for a range of initial concentrations of hexyne and the catalysts. When mercuric acetate and perchloric acid were used as the catalysts , the initial rates were found to be interpretable on the assumption that the reactive species was a 1 to 1 to 1 complex formed by the 3-hexyne, the mercuric acetate, and the perchloric acid. By means of indicator studies the existence of the 1 to 1 to 1 complex (probably a hexyne-mercuric monoacetate cation) was proved and its equilibrium constant measured. The indicator studies were extended to the measurement of the basicities of weak bases in acetic acid. A simple relationship was found between the observed ionization constant and the thermodynamic activity constant. New values were found for the basic strengths of dioxane, acetonitrile, p-toluene-sulfonamide, water, acetanilide, acetamide, and urea. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry and Physics) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Minor Option: Physics Thesis Availability: Public (worldwide access) Research Advisor(s): Lucas, Howard J. Thesis Committee: Unknown, Unknown Defense Date: 1950 Record Number: CaltechTHESIS:10312023-185511760 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:10312023-185511760 DOI: 10.7907/5g6h-b137 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16224 Collection: CaltechTHESIS Deposited By: Kathy Johnson Deposited On: 31 Oct 2023 19:04 Last Modified: 31 Oct 2023 19:18 Thesis Files PDF - Final Version See Usage Policy. 51MB Repository Staff Only: item control page

ADDITION TO THE CARBON-CARBON TRIPLE BOND THE REACTION OF ACETIC ACID WITH 3-HEXYNE Thesis by Henry Lemaire In Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1950 Acknowled gments It is a p le a sure t o ac k nowl e d ge my indebtedn ess to Profess or H. J. Lucas. He proposed the problem on which this thesis is based, helped g uide the research through a considerable number of difficulties, and is, to a lar ge extent, responsible for t h e c ompletion of the wor k . I wo uld als o like t o r e cord my g ratitude to sever a l collea gues, Mr. W. Smith Dorsey and Mr~ Nathan Koenig in particular, for much valuable assistance. Abstract The mechanis m of the mercury-catalyzed addition to the carbon-c a rbon triple bond has been inve sti gated. was used as the solvent and one reactant. was 3-hexyne. Acetic acid The other reactant The course of the reaction was follo wed by bromine titrati on and by a sp ecial indicator method. The addition was found t o take place in two discrete steps, forming 3-acetoxyhexene-3 and 3-hexanone. The rate of the i nitial reaction was determined for a ran ge of initial concentrations of hexyne and the catalysts. When mercuric acetate and perchloric acid were used as the catalyst s , the initial rates were found t o be interpret a ble on the asswnption that the reactive species was a 1 to 1 to 1 complex f or med b y the 3-hexyne, the mercuric acetate, and the perchloric acid. By me ans of indicator studies the existence of the 1 to 1 to 1 complex (probably a hexyne-mercuric monoacetate cation) was p roved and its equilibrium constant measured. The indicator studies were extended to the measurement of the basicities of we a k bases in acetic acid. A simp le rel a tion- ship was found between the observe d ionizatio n con stant and the thermodynamic activit y c onstant. New values were found for the basic strengths of dioxane, acetonitrile, p-toluenesulfonamide, water, acetanilide, acetamide, and urea. Table of Contents PART TITLE PAGE I. Historical Background and Discussion of Problem . • . . . . . . . . . . 1 Preparation cf Materials and Isolation of Products. . • . . . . . . . . . . 6 .. 25 II. III. IV. Methods of Analysis . . A. Analysis of 3-Hexyne . . . 25 B. Determination of Hexenyl Acetate 33 c. The Bromination of Dtethyl Ketone . . 37 D. The Action of Diethyl Carbitol in the Bromide-Bromate Titration . • . • 38 Reaction of 3-Hexyne and 3-Acetoxyhexene-3 with Acetic Acid. Preliminary Results • 40 A. The Reaction of 3-Hexyne 40 1. Calculations . . • . • 40 2. Reactions catalyzed by perchloric acid and mercuric acetate. Preliminary Results . . . . . . . . . . . . . . 42 Reactions catalyzed by boron trifluoride and mercuric acetate . • . . • . . 46 Reactions catalyzed by sulfuric acid and mercuric acetate . . • . . . . 46 Studies relating to the yellow color appearing in solutions of hexyne, mercuric acetate, and perchloric acid . . . . . . . . . . . . . . 54 3. 4. 5. B. The Reaction of 3-Hexenyl Acetate-3 with Acetic Acid Catalyzed by Mercuric Acetate and Boron Trifluoride . . . 70 Table of Contents (Continued) PART TITLE PAGE v. The Measurement of Basicity in Glacial Acetic . Acid . . . . . . . . . . . . . . . . . . 72 A. B. Preliminary Investigations. with Sulfuric Acid. . . . . . . . . . 72 Extended Investigations with Perchloric Acid. . . . . . . . . . 76 1. Introduction and theory of calculation. 76 2. Detailed method of calculation. limitations of the method. 82 The preparation of indicators and the properties of their solutions in acetic acid. • • • • • • . • • . . 87 The purification of the colorless bases and the determination of their basic stren g ths • • . • • . . • . 98 c. The Formation of Complexes with Mercuric Acetate by Hexyne and Hexenyl Acetate . 104 D. The Acidity of Solutions of Sulfuric Acid in Acetic Acid . • • • • • . . • . . • • 111 Reaction of 3-Hexyne and 3-Acetoxyhexene-3 with Acetic Acid . . . . . . . . . . . . . . . . . 113 3. 4. VI. A. B. . .. . Special Tests • • • • . . . . . . . . . . . . The Mechanism of the Reaction . • 113 126 c. The Co r respondence of the Amount of Yellow By-product to the Loss of Unsaturation. 128 D. The Rate of the Reaction of the HexyneMercuric Ac e tate-Perchloric Acid Co mp lex with Acetic Acid . . . . . . 131 The Rate of Reaction of 3-Acet oxyhexene-3 with Acetic Acid. Catalysis by Mercuric J c e tate and Perchloric Acid . . 134 . . . . . . . . . . . . . . . . . . . . 138 E. VII. The Summary • I. Historical Background and Discussion of Problem In 1936, Henn ion, Vogt, and Nieuwland published "A Proposed Mechanism for Mercury Catalysis in Acetylene Addition Reactions" 1 • Actually, two different mechanisms were suggested for the reaction of hydroxyl compounds with ~lkynes. One proposa:J,. - was that the mercuric salt reacted with the alkyne to form a metalated alkene and the other was that the mercuric salt reacted first with the hydroxyl compound. The first mechanism is shown in Equations 1, 2, and 3 and the second, in Equations 4 and 5. HgA * I HgA2 • R-CiC-R --> ➔ R-C:C-R (1) I A HgA BO HgA I * l I ♦ BOH -HA R-C-C-R in one or two steps-,, R-CaC-R ,- (2) A BO HgA I I R-C:C-R +HA -HgA2 ► R-C:CHR I (3) OB HgA 2 + BOH -HA > BO - HgA (4) *Bo BO - HgA. R-c;c-R * Cis or trans compoundsY HgA I I R-C = C-R (5) -2- The product of Reactions 3 and 5 would either rearrange to form a ketone (if B was H) or would add another mole of BOH without specific catalysis (if B was an alkyl group).# The next year Frieman, Kennedy, and Lucas published 3 the results of a study of "The Rate of Hydration of Acetylene in Aqueous Solution of Sulfuric acid and Mercuric Sulfste". They had studied the formation of acetaldehyde from acetylene and had found the initial rate to be proportional to the first power of the acetylene concentration and the second power of the mercuric sulfate concentration. The dependence of the rate upon the concentration of sulfuric acid was not determined. Classical studies on the complexes formed between unsaturated organic compounds and mercuric salts were made by J. Sand and several collaborators. Sand and Breest 4 studied the equilibrium/which they wrote: (6) by measuring the conductivity of the aqueous solution. Hugel 5 and Hibou found that mercuric acetate formed complexes with olefins in glacial acid of the formula C0 H2n•Hg(OAc) 2 • On the addition of water these complexes could be obtained as oils which crystallized rapidly. One of the acetate radicals # On the basis of experimental work, Wright 2 had concluded that reactions an a.lo.go us to Equations 4 and 5 should be preferred in explaining the formation of olefin complexes. -3- was found to be very readily replaced by an hydroxyl radical on titration with alkali. In ether, the second acetate could be replaced slowly by alkali. Much interest has been centered on the question of whether the products of the reactions between olefines and mercuric salts are loose complexes or saturated, mercurated compounds of a somewhat unstable nature 6 • For the purposes of this paper it is important to note that uponaddition of mineral acid (as in a bromide-bromate titration) the olefin is completely regenerated. Lucas, Hepner, and Winstein 7 measured the equi- librium of cyclohexene and mercuric nitrate in aqueous solution. Their data indicated the occurrence of two equilibria; B) (Hgt ♦ ) • 2.2 X 104 (8) ~B• HgOH") (H") = 5. 0 X 10 4 (9) B) (Hg-1-1) where Bis cyclohexene. The purpose of the work described in this thesis was to acquire new data on the mercury catalyzed addition to the carbon-carbon triple bond. In order to avoid any complications from the acidic nature of the ,hydrogen atoms of acetylene, diethylacetylene was chosen for investigation. After some preliminary work, acetic acid was chosen as solvent and second reactant. This decision was based on both theoretical and practical considerations. When water is added to the triple bond, the initial product, a vinyl alcohol, rearranges spontaneously to form an aldehyde or ketone. It is, therefore, not possible to study the addition reaction separately. Acetic acid, on the other hand, adds to alkynes to form stable derivatives of vinyl acetate. A second addition was believed to take place, but this reaction could be studied separately and its effect on the initial reaction determined. Acetic acid will dissolve hexyne, which is but sparingly soluble in water. Moreover, acetic acid has a high enough dielectric constant so that the mercuric salt catalysts were expected to be sufficiently soluble in it to make possible a homogeneous system. Finally, acetic acid was known to be stable to bromination, the analytical method which seemed best adapted for following the rate of reaction. The course of the reaction between alkynes and acetic acid was investigated by Hennion and Nieuwland 8 • They reacted acetic acid with propyne, 1-hexyne and 1-heptyne. The catalyst was prepared by dissolving mercuric oxide in a mixture of methanol and ether-boron trifluoride and then in the acetic acid. The alkyne was added dropwise. Two products were isolated, the alkylvinyl acetate and the ketone. R-C:CH • HOAc -----"►- R-C:CH2 I OAc (10} R-C:C¾ + HOAc --->;;,. I OAc R-C-CH3 + Ac20 (11) II 0 The ester was obtained in 30% - 34% yield. It was not found possible to isolate a ketone diester, and the authors were not certain whether the ketone was formed by Reaction 11 or by hydrolysis of the mono-ester during isolation. 1-Hexenylacetate~2 boiled 74-47 o at 39 mm. and had nD26 : 1.4176. -6- II. Preparation of Materials and Isolation of Products 3-Hexyne. A procedure published by Bried and Hennion was adapted for the synthesis of 3-hexyne. 9 Sodium acetylide is prepared and, in the · presence of sodamide, reacted with ethyl bromide to form butyne-1. This alkyne is converted to its sodium salt by the sodamide and then reacts with a second mole of ·ethyl bromide to form 3-hexyne. The four successive steps are shown below. CH:CH + Na CH:CNa (12) CH:C-Et + NaBr (13) CH:C-Et + NaNH 2 NaC::C-Et + NH3 (14) NaC:C-Et + EtBr ---~- Et-C:C-Et + NaBr (15) Sodium acetylide was prepared in liquid ammonia by adding to a reaction flask containing ammonia a stream of acetylene and a solution of sodium in ammonia. (The acetylene should be kept flowing continuously - a point on which the literature 10 is not clear and should always be in excess over the sodium.) Three moles of sodium were dissolved in 1.5 liters of ammonia (which had to be replenished during the run) and forc·ed over by the pressure of the ammonia into the 5-liter reaction flask. Ammonia causing excess pressure was allowed to escape from the sodium solution through an oil column. It was found that 6 mm. pyrex tubing and 3/16 11 rubber tubing became clogged with sodium during the periods when it was necessary to stop -7- the flow of the sodium solution. Eight mm. pyrex tubing and 1/4 11 rubber tubing were satisfactory. Acetylene was led from a tank through aqueous sodium bisulfite, calcium chloride, and soda lime into the reaction flask, which was initially charged with about 2 liters of ammonia arid fitted with a dry ice condenser. Vigorous stirring with a half-moon stirrer was employed and the _proportion of entering acetylene rand sodium solution so regulated that . the entire reacting solution never became entirely blue. Under optimum conditions such a run could be completed intwo and a half hours. Sodamide was prepared in liquid ammonia by dissolving about a tenth of a gram of ferric nitrate nonahydrate in a liter and a half of ammonia. When the entire solution had become orange about a gram of sodium was added and air, dried with potassium hydroxide, bubbled through until the blue color disappeared entirely, giving way to a bleck suspension. Four moles of sodium were then added and time allowed for the reaction to be completed (1 to 3 hours). Very little trouble was en~ countered in this preparation. For the preparation of 3-hexyne the sodamide slurry had to be transferred to the 5-liter reaction flask. Considerable difficulty was encountered in effecting this transfer through glass and rubber tubing, and in the three most successful runs the slurry was simply poured. (It is notable that no great effort had to be taken in this series of preparations to keep out atmospheric moisture.} To the mixture of sodium acetylide and sodamide 5 moles of ethyl bromide (dried with calcium chloride and distilled) were added dropwise with vigorous stirring over a period of two to three hours and water was then added until the reaction flask contained clearly defined aqueous and organic phases. During the addition of water a copious stream of ~m.~onia WQS evolved which was led over water in a filter flask, but in no run did & separate phase ever appear i n this trap. The organic layer was washed thoroughly with water, dilute hydrochloric acid, water, dilute sodium carbonate, and water. The crude yields in f our runs were 54 g., 135 g., 106 g., and 130 g. When these lots were com- bined and fractionated, a yield of 317 g. of pure 3-hexyne (Nfi 50 : 1.4082) was obtained. best run was therefore, The indicated yield on the g~ .l x 317 = 1.2 moles or 48%. Sodium 425 5 acetylide is the limiting reagent. Hexyne is unstable in air, reacting with oxygen to form peroxides. Some commercial 3-hexyne, purchased from Farchan, was found to have so much peroxide in it that the potassium thiocyanate - ferrous ammonium sulfate solution with which it was shaken became quite warm, and many extractions were necesw sary to remove the peroxide. On the other hand some hexyne which was rinsed outafacolwnn with acetone and kept in this solvent for three years in a loosely corked flask was found to be nearly free of peroxides. storage of hexyne. Acetone is recommended for long -9- Three values for the melting point of 3-hexyne are / 11 012 013 available in recent literature, -51° , -101 , and -105.5 • Several efforts to freeze the freshly distilled hexyne in an acetone bath at 78° were .unsuccessful. The material froze in liquid air but melted again in the Dry Ice bath. In the many distillations of 3-hexyne made during the course of this work, the major portion of the product always boiled between 79.9&80.2° at atmospheric pressure (about 745 mm.) ·: Most of these distillations were made under nitrogen. In one such distillation (at 742.8 mm.) the main fraction almost all distilled at 80.0°. The same thermometer was then employed to measure the boiling point of C.P. benzene and was found to read 78.8°. If the stem and barometric corrections are assumed to be additive over the small range involved, the corrected boiling point of 3-hexyne is 80.0 • (80.1 - 78.8): 81.3° ~t 12 760 mm. Campbell and Eby give 01.2° - 81.3° at 747 mm., whereas the later paper of Henne and Greenlee 13 , devoted ex• elusively to physical constants, gives merely 81° at 760 mm. Acetic acid. Two methods were used for the prep aration of anhydrous acetic acid: refluxing with acetic anhydride and fractional distillation. Fractional crystallization was attempted in a preliminary test . but was found to be wasteful of acetic acid,and two recrystallizations raised the melting point only to 16.0°. Baker's Chemical Reagent acetic acid melts, usually, at 15.6° while pure acetic acid melts at 16 . 6°. Treatment with -10- acetic anhydride was mo r e efficaceous. The meltin g point of the rea g ent a cid was taken and t h e amount of water present was calculated 14 on the assumption that the only impurity responsi b le for the freezing point de pression was water. The calculated amount of acetic anhydride was then added and the liquid was refluxed for two to three hcur~~ In this way various lots, melting from 16.2° to 16.5°, were obtained. Then it was found (the first time accidentally) that o verni ght refluxing gave ' a product with a me lting point of 16.6°. Most of the early kinetic runs we r e made in acetic acid which had been refluxed overnight with about 0.05 M excess acetic anhydride. of the runs extra acetic anhy dride was added. To some It was believed, and subsequent work clearly i n dicated, that small amounts of eithe r water or acetic a n h ydride would not g reatly effect the rate of the reaction, but th a t acetic anhydride was to be preferred to water. The acetic acid for all indicat or wor k and for the concludin g and conclusive kibetic runs was prepared by fractionation through a 90 cm. column of g lass helices fitted with an electrically heated jacket and a total-reflux partial-takeoff still head. 15 Batch lota of 2 1/2 liters were distilled and initial cuts of 500-700 ml. were discarded. alway s melted at 16.6°. The main product The various batches of purified acetic acid were kept in flasks fitted with a 24/40 standard-taper stopper. The acetic acid ·,,; as found to be quite stable during continual use. In some cases melting points ta k en weeks apart showed no chan g e whatever. -11- The thermometer used for taking these melting points was calibrated in 0.2°, and readings which coincided with a calibration were probably accurate to 0.02°. The melting point of one batch of acetic acid was taken with both the usual laboratory thermometer and with N.B.S. thermometer #42940, calibrated in 0.1°. Both thermometers read 16.60°. The Bureau of Standards certificate gave the following data. Thermometer Reading True Temperature 10.00° 10.03° 20.00° 19.98° It is, therefore, concluded that the laboratory thermometer gives the true temperature to within 0.02° in the vicinity of 16.6°. Cyclohexene. This alkene was prepared by fractionating 100 ml. of the Eastman product through 30 cm. of glass helices. A 20 ml. central portion was taken over a 0.2° range. Carbon tetrachloride. This solvent was purified by saturat- ing 2 1/2 liters of material with chlorine gas and exposing to sunlight for a week. The solution was then washed with aqueous sodium hydroxide, dried over calcium chloride for a day, and distilled through a Vigreux column. distillate was cloudy. The first 200 ml. of The main portion, more than 2 liters, was stored in glass stoppered bottles. Diethyl carbitol (bis P-ethoxyethyl ether}. Carbide and Carbon diethyl "cellosolve" solvent was stored for a week over -12- potassium hydroxide and then distilled at .68-70 mm. The first 300 ml. were discarded and about 1 liter was collected and stored under nitrogen. A few ml. were heated in a test tube with potassium hydroxide. A second phase developed, but only a faint trace of color was visible. To another sample of the ether one drop of 2% aqueous potassium permanganate was added. There was no evidence of reaction. Finally, to an aqueous solution of ferrous ammonium. sulfate and ammonium thiocyanate a small amount of the carbitol was added. No color developed. It is, therefore, concluded that the solvent was free of both aldehydes and peroxides. Mercuric salts. General Chemical Company, C.P. mercuric acetate was used without purification. Its solubility in glacial acetic acid was found to be just short of 0.100 M. Such saturated solutions were made up as stock solutions but were always diluted to give an intermediate standard solution before being used in a kinetic or . colorimetric determination. Reagent grade mercuric sulfate was used in an attempt to make up a standard stock solution in acetic acid, but the salt was either so slightly or so slowly soluble that a solution 0.0002 M could not be obtained. Use was made of mercuric sulfate as a catalyst by mixing acetic acid solutions of mercuric acetate and sulfuric acid in the presence of hexyne. Although some trouble was experienced due to precipi- tation ; much data was thus obtained. Mercuric perchlorate was prepared by dissolving 0.46 mols of C. P. red me.rcuric oxide in O. 972 mols of HC10 4 in 60% aqueous -13- HCl0 . After several hours a still undissolved suspension was 4 filtered off. After standing overnight at 5° in the cold room the crystals of mercuric perchlorate were filtered off with suction, bottled, and left in the cold room. cent, however, and were not used. They were deliques- Mercuric perchlorate, like mercuric sulfate, was prepared in solution as needed by mixing acetic acid solutions of mercuric acetate and perchloric acid. Strong Acids. Perchloric acid was prepared from a 60% aqueous solution of Baker's C.P. reagent. The normality of the aqueous solution was determined by diluting an aliquot from a total-delivery pipette 16 to provide about a 0.10 N aqueous solution, the stren gth of which was accurately determined by titration against an iodide-iodate solution 17 • The 60% solution was found to be 9.011 Nin perchloric acid. A solution of this strength has 905.4 g. of acid per liter and a specific 18 gravity of 1.529 • A ten-ml. volumetric pipette delivered 15.27 g. of the strong acid. Mr. Cunningham had found a specific gravity of 1.533 for the same solution. The value 1.530 was used to calculate that the soluti on was 59.16% HC10 4 and 40.84% HzO. Bak er's label gave an assay of 59.1% HC10 4 • Acetic acid solutions cf perchloric acid were prepared in strengths ranging from 0.525 M to 0.0500 M by mixing in - - anhydrous acetic acid the appropriate quantities of acetic anhydride and aqueous perchloric acid. In the presence of the mineral acid the reaction between the water and acetic anhydride took place quite rapidly with the evolution of considerable heat. The solutions were adjusted to final volume -14- with acetic acid after the mixtures had returned to room temperature. More dilute stock solutions of perchloric acid in acetic acid were prepared by diluting these solutions with glacial acetic acid. An 0.0500 M solution prepared according to these calculations was titrated with an acetic acid solution of freshly distilled aniline with crystal violet as indicator. The solution was found to be 0.0500 Min perchloric acid. An acetic acid solution 0.0500 Min mercuric acetate and 0.200 Min perchloric acid was prepared by mixing the appropriate solutions. This solution was found to be quite stable. However, a solution prepared so as to be 0.100 Min mercuric acetate and 0.200 Min perchloric acid formed a precipitate, which was very slightly soluble even in water. An attempt to prepare a solution half this strength - 0.0500 Min mercuric acetate and 0.100 Min perchloric acid - also resulted in a water insoluble precipitate. Boron trifluoride etherate, generously supplied by the General Chemical Company, was received as a black liquid which fumed vigorously in air. Distillation yielded a nearly color- less material, boiling point range 123.0° - 123.5°. A solution 0.200 Min boron trifluoride was prepared by making 25.0 ml. of the distillate up to a liter with acetic acid. The solution seemed to be fairly stable, but the unused distilled material became noticeably darker in one day. An acetic acid solution 0.050 Min the hypothetical compound Hg{0Ac)z•2BF 3 was found to yield a precipitate but -15- solutions from 0.00100 M to 0.0100 M were stable. occasion On another a solution 0.0200 Min boron trifluoride and 0.00500 M in mercuric acetate gave a precipitate on standing for two days. The solid was not soluble in water, though it seemed to react with water to form a new precipitate. A precipitate was observed after three hours in a solution 0.0200 Min boron trifluoride and 0.00300 Min mercuric acetate. Precipitates, perhaps of a different nature, were also formed occasibna11r in the presence of hexyne. Sulfuric acid with a melting point of about a0 was pre- pared by mixing 50 ml. of 25% fuming C.P. sulfuric acid with 100 ml. of reagent grade (96%) sulfuric acid. Solutions in acetic acid were prepared from weighed amounts of this material. 3-Acetoxyhexene-3. Four attempts were made initially to prepare 3-acetoxyhexene-3 (also called hexenyl acetate and diethyl vinyl acetate) from hexyne with mercuric acetate and perchloric acid in acetic acid. shown in Table 1. The details of the runs are The s ynthesis being attempted is shown in Equation 16. Table 1. Attempted Preparation of Hexenyl Acetate. Run 359 361 365 388 Hexyne moles/1. Hg(OAc}2 moles/1. HCl04 moles/1. 1.1 0.51 0.48 0.50 0.010 0.005 0.005 0.010 0.038 0.020 0.020 0.020 -16- Et-C = C-Et. Et-CiC-Et • HOAc ' AcO (16) I H Unexpected colors appeared in these runs, ranging from pale yellow to deep red and brown. in each. Precipitates were also formed No hexenyl acetate could be isolated, and, in the case of the third and fourth runs, it was found that only a very small fraction of the unsaturation disappeared in 24 hours. Since Hennion and Nieuwland accomplished the synthesis of 2-acetoxyhexene-l from 1-hexyne with boron trifluoride, this acid was employed next. One hundred ml. of acetic acid solution 0.35 Min hexyne, (3.5 gm. of hexyne) 0.0200 Min boron trifluoride, and 0.00500 Min mercuric acetate (a similar run with 0.00100 M mercuric acetate gave no product) were allowed to stand at 25° for 24 hours. Titration then indicated the loss of half the original unsaturation and the 90 ml. of solution were steam distilled from 600 ml. of boiling water. The first 30 ml. of distillate included four ml. of an organic phase which was dried over calcium sulfate and distilled at 40 mm. A product weighing 2.4 g. was collected at 67-70°. Allowing for the aliquots taken for titration during the run, the yield was 55%. The steam distillation was continued until about 150 ml. of aqueous solution was collected. This material was neutralized with potassium carbonate and extracted with ether, which was then dried with calcium chloride. After stripping off the ether a residue of about 1/2 ml. was obtained. -17- Four such runs were made, two on a somewhat larger scale, with the products of each run being added to the crude material of the succeeding run for distillation. The final distillation was made at 40 mm. and six fractions collected as shown in Table 2. Table 2. · Distills tion of hexenyl acetate at 40 mm. Fraction No. Temp. 0 Weight (g. ) !!n 1.9 1.4188 1 69.5°-72.a 2 72.8° 2.4 1.4198 3 72.8° 2.1 1.4197 4 72.8° 2.6 1.4197 5 72.8° 2.4 1.4196 6 72.8°-72.9° 2.5 1.4196 The temperature control on the refractometer was inadequate when the readings were taken, the temperature varying 25° 26°. Fraction 4 was submitted for a carbon hydrogen analysis. Calc'd. for c 8 H1402: C, 67.57%; H, 9.93%; found, C, 68.42%; H, 10.26%. Bromine titrations of aliquots containing 1.644 milliequivalents g~ve titers of 1.63, 1.648, and 1.653 milliequivalents. Fraction No. 5 was used for a density determination. A pycnometer, kindly loaned by Dr. Harold Garner, with a volume of 1.6814 ml. at 25°, was found to hold 1.4924 g. of hexenyl acetate at 25°. Correction for the displaced air gives -1819 The specific gravity, a ca 1 cu1 a t e d mass o f 1.4942 g. 25 d 25 is, then, 0.888. The molar refraction was calculated from these data and found to be 40.49. The theoretical molar refraction for hexenyl acetate is 40.33. 20 As in the runs with perchloric acid, the solutions of hexyne, boron trifluoride, and mercuric acetate gave colors and, usually, precipitates. The colors observed ranged from yellow through red to brown. In the absence of mercury, a boron trifluoride-hexyne solution became a quite pretty rose color without any appreciable loss of unsaturation. A determined attack was then made on the problem of synthesizing hexenyl acetate with perchloric acid as the acid catalyst. Nineteen different solutions were prepared, as shown in Table 3, and tested at intervals for loss of unsaturation. -19- Table 3. Attempted preparations of hexenyl acetate with perchloric acid. Volumes 30 to 50 ml. Hexyne Hg(0Ac) 2 HCl04 0.34 0.0050 0.0200 2a 0.42 0.0238 0.0485 3a 0.29 0.0161 0.0323 4a 0.29 0.0242 0.0161 5a 0.25 0.0278 0.0139 6a 0.22 0.0122 0.0244 7a 0.25 0.0278 0.0139 Sa 0.02 0.0125 0.0050 9a 0.11 0.0125 0.0050 lla .04 0.0125 0.0050 12a .07 0.0125 0.0050 139. • 09 0.0125 0.0050 14a 0.11 0.0077 0.0077 15a 0.35 0.0050 0.0200 16a 0.35 0.0030 0.0200 17a 0.18 0.0030 0.0200 18a 0.11 0.0037 0.0250 G 19a 0.088 0.0050 0.0300 G Run M la M M G Runs 8, 18, a n d 19 were the on ly ones t o s h ow sufficient loss of unsaturation to give promise of yielding much hexenyl acetate. -20- A large scale run using the same concentrations of materials as were used in Run 19 was then made. Ten ml. of hexyne were dissolved in a liter of catalyst solution. After four and a half hours the unsaturation had dropped from 0.35 to 0.24 equivalent weights per liter. The one liter of solution was steam distilled from three liters of boiling water. The acetic acid was added gradually during the distillation, and it was noted that towards the end of the run, when the still pot residue was 15%-25% acetic acid, much less organic phase was distilled. (It should be noted that the ratio of water to acetic acid is less than half the corresponding value in the boron trifluoride runs. This fact has an important bearing on the lower yields obtained with perchloric acid.) About 250 ml. of distillate was collected and extracted with ether. The ether solution was dried (calcium chloride seemed to be better than calcium sulfate for this purpose) and the ether stripped off at atmospheric pressure. The products from four such runs were combined and distilled at 41 mm. pressure. 68-78°. Six and a half ml. were collected from The dry ice trap contained water and about 8 ml. of organic material which smelled strongly of hexenyl acetate. The 6 1/2 ml. were combin.ed with 1 1/2 ml. from a previous similar (single) run and 3 ml. from the boron trifluoride synthesis (Table 2) and redistilled. At 31 mm. 7 ml. distilled from 63°-69° and ... 21- at 25 mm. 3 1/2 ml. distilled from 68°-71°. 25 The indices, nD were 1.4239 and 1.4231 respectively at 25.0. These products were recombined and fractionated through a semi-micro column with a coiled glass rod for packing. At 33-34 mm. four fractions were collected. Table 4. Distillation of hexenyl acetate at 33-34 mm. 25.0° Temp. Volume (ml.) !1n A 52°-59° 0.6 1.4180 B 61. 5°-66° 2.0 1.4260 C 69°-74° 5.5 1.4229 D 73o_75o 1.1 1.4212 Fraction No. Fraction B was later redistilled and collected in quite good n25 equal to 1.4233, so it may be that the high Yield with -D reading for B above was due to impurity in the apparatus. It is apparent, however, both from the boiling point range and from the index of refraction that this product differs from that obtained with boron trifluoride. was probably about 15%. The best yield Bromine titration indicates a saturated impurity amounting to about 6%, and the analysis for carbon was low. Calculated for c8 H140 2 : C, 67.6%; H, 9.9%; found: C, 65.9%; H, 9.9%. 3-Hexanone. 3-Hexanone was prepared from 3-hexyne in an acetic acid solution 0.25 Min hexyne and a little less than -22- 0.020 Min boron trifluoride and 0.010 Min mercuric acetate. (The reaction mixture contained 20% of the supernatant liquid from a solution 0.050 Min mercuric acetate and 0.100 Min boron trifluoride. A slight precipitate had deposited in this solution.) After 29 hours 93% of the unsaturation had disappeared. The 80 ml. of reaction mixture was steam distilled from 200 ml. ~ of water • . About 1.9 ml. of organic layer were collected in 25 ml. of distillate. This layer was dried with calcium chloride and calcium sulfate and distilled with tert.-butylbenzene as a still base. About 0.3 g. of product distilled at 118°-124° and an equal amount was obtained at 124°-124.5°. The calculated The first cut had n5 5 equal to 1.4000. 21 22 The literature gives a boiling potnt of 123° and !!n . equal 22 to 1.3990 • The 2,4-dinitrophenylhydrazones of some ketones 25 are di- or polymorphic • The melting point of the hexanone-3 24 25 derivative has been reported as 130° and 146.5°-148.so • yield was about 30%. The above two 0.3 g. cuts were mixed and used to make the 24 derivative according to Allen's directions. RecrystallizaM tion from ethyl alcohol and acetic acid gave a product melting 127°-129°. The reaction between hexyne and acetic acid to yield 3-hexanone probably takes place in two steps. The hexyne reacts first with one molecule of acetic acid to give 3acetoxyhexene-3 (Equation 16). A second molecule of acetic acid then adds to this ester to form an unstable intermediate -23- which, as shown in Equation 17, breaks down to give 3-hexanone and acetic anhydride. AcO f l Et-C: CH-Et+HOAc --~Et-C-CH2-Et ,, 0 AcO ' -->--. Et-C-C¾-Et + Ac20 (17) AcO Because they were unable to isolate the ketone diester in . certain analogous reactions, Hennion and Nieuwland 8 were uncertain whether Reaction 17 actually took place or whether the ketones they isolated were formed by hydrolysis of the substituted vinyl esters during the isolation. The nearly complete loss of unsaturation ob~erved in the above synthesis of 3-hexanone shows that Reaction 17 actually took place. Tentative evidence will be presented later that the ketone di ester is unstable, decomposing even in the anhydrous s.o lution. Analytical reagents. Merck reagent grade potassium bromate·was used as the primary standard for the analytical determinations. It was dried in an oven at 110° for three days and thereafter stored in a vacuum desiccator over calcium chloride and phosphorus pentoxide. Two-liter batches of 0.1000 N bromide-bromate solution were made up with 5.568 g. (l/30 mole) of the potassium bromate. During the early part of the work 35 to 40 g. of sodium or potassium bromide (about 1/3 mole) were used in preparing the solutions, but later, at Mr. Dorsey's suggestion, the amount of bromide was reduced to 1/6 mole. Solutions of sodium thiosulfate (0.05 M) were -24- prepared by dissolving 25 g. of Baker's C.P. pentahydrate and about two grams of sodium carbonate in 2 liters of water. This solution was standardized against the bromate solution with the same technique as was being employed for the hexyne determinations. The need for running blanks was thus avoided. Occas16nallt · the thiosulfate solutions were also standardized with potassium dichromate 28 and the results always agreed well with the bromate standardizations. were found to be stable. The thiosulfate solutions Some solutions showed no change (0. 05,00 li to 0. 0500 N for example) in two months while one, perhaps the least stable, changed from 0.0505 to 0.0501 in 40 days. Because the equivalents of thiosulfate used in a titration were usually about 25% of the equivalents of hexyne present and rarely more than 50%; the normality· of the thio• sulfate did not have to be known to an accuracy greater than about 4 parts in 1000. Mercuric sulfate solutions were made up in 6 ! sulfuric acid. The 0.2 M solutions presented no difficulties in their preparation. ring. The 0.6 !! solution s required heating and stir- Potassium iodide solutions were made up in small quantity, to last only two or three weeks. The pure solutions slowly turn ed yellow, but since this color - even in 200-300 ml. of solution - could be discharged with a very few drops of 0.05 N thiosulfate solution, it was not considered detrimental. Potassium iodide solutions made up with a gram or two of potassium carbonate were quite stable. -25- III. Methods of Analysis A. Analysis of 3-Hexyne The .carbon-carbon triple bond does not add bromine quantitatively in a simple bromate titration; but in the presence of mercuric sulfate a smooth, rapid addition can be secured 26 • It was found necessary to use a quantity of mercuric sulfate in excess of the total halide present during the bromination and then to add excess sodium chloride before the final iodide-thiosulfate titration. This procedure was further studied by Lucas and Pressman27 , whose directions are herewith summarized. A 10 to 15% excess of 0.1 N bromide-bromate solution is run from a burette into-a conical flask having a ground-glass stopper bearing a sealed-in stopcock. Following the evacuation of the flask, 5 ml. of 6 N sulfuric acid are added and allowed to stand 2 to 3 minutes. Next, 10 to 20 ml. of 0.2 M (not 0.2 N the correction is from a private communication from Dr. Pressman) mercuric sulfate and the solution to be analyzed, which should have about two milliequivalents of unsaturation, are run in. About 15 ml. of carbon tetrachloride were used for washing in the sample, which was a carbon tetrachloride solution of the unsaturated compound, and 20 ml. of acetic acid were added. The flask, wrapped in black cloth, was shaken for 7 minutes, and 15 ml. of 2 N sodium chloride and 15 ml. of 20% potassium iodide were then added. After 0.5 minutes final shaking, titration was made with 0.05 N sodium thiosulfate, using a starch solution as indicator. The accuracy of this method varied with the different alkynes used, ranging from +2.2% for 1-pentyne to w3.5% for 2-heptyne. 27 (No acetic acid was used in the pentyne analyses.) Average deviations ranged up to 1.4%. In the course of the -26- present work several changes were made in the above procedure, and they will be discussed in this section. Some of these chan ges, as noted, were suggested by Mr. W. Smith Dorsey, who has made an extensive investigation of the method. It is evident that an average deviation of 1.4%, representing, as it probably does, a dispersion of about 5%, is not satisfactory for kinetic studies. When hexyne reacts with acetic acid to form hexenyl acetate, the am ount cf unsaturation chan g es. It is not possible to make the initial analysis recommended by Lucas and Pressman. In order to arrive at the suggested 10-15% excess of bromate over alkyne, it was necessary to guess at the amount of unsaturation present in successive aliquots. Much work was wasted and time lost before it wis realized that about 15% excess bromate should be not a maximum but a minimum requirement. If less bromate than this excess is used, some of the 'hexyne will remain unreacted; and an apparent, but spurious, 10-15% excess bromine will be found. Therefore, by guessing that a reaction will take place and adjusting the bromate titers accordingly, it is possible to make the reaction seem to take place. By maintainin g the concentration of bromate at least 15-20% in excess of the hexyne, this effect was eliminated. It was found that an excess of bromate up to 55% had little effect on the quantity of hexyne found. data are shown in Table 5. The -27- Table 5. The Effect of excess Bromate Solution used in the Titration of Successive Aliquot s of ap Acetic Acid Solution of 3-Hexyne containing 1.50 ,Milliequivalents of Hexyne. (Analysis Method II) (Quantities are in milliequivalents.) Titration No. Time for Bromination (Min.) Bromate Used Hexyne Found 1 5 1.60 1.47 2 5 1.80 1.58 3 5 2.00 1.60 4 5 2.22 1. 60, 5 5 2.50 l.6l. 6 15 2.50 1.63 Titration No. 1 of Table 5 is an example of the fallacious results which are obtained when not enough bromate is present. The results of titrations 2-5 show an average deviation of ,0.5%, as the excess bromate ranges from 14% to 55% of the hexyne found. In titration No. 6 the hexyne was left in contact with the large excess of bromine for 15 minutes before sodium chloride and potassium iodide were added and the solution was back-titrated with thiosulfate solution. It is' of interest that even in these extreme conditions the hexyne found was only 1.4% above the mean of titrations 2-5. As recommended by Lucas and Pressman, 5 ml. of 6 N sulfuric acid were added to the bromide-bromate solution. However, a minimum of five minutes were always provided for the liberation of the bromine. -28- Samples of the reaction solutions were taken in a pipette calibrated for total delivery16 • Sampling was effected through an all-glass double stopcock assembly which allowed the sample to be forced up into the pipette by the pressure of nitrogen from a· cylinder. Nothing but the nitro-- gen and pyrex glass came into contact with the solution. The sampling apparatus is shown in Figure 1. The nitrogen used for most of the work was water-pumped, 99.5% pure material dried by passage through a tower of calcium chloride. the conclusive For kinetic runs, nitrogen with a minimum purity of 99.99% was employed. During some preliminary work using a carbon tetrachloride solution of cyclohexene, an attempt was made to force the sample up with mercury. It was found that the metal reacted with the cyclohexene in solution! The size of the aliquot of hexyne solution which it was convenient to take depended on the concentration of the hexyne. A list of the total delivery pipettes and their calibrations is shown in Table 6. In the present work the analytical solutions were homogeneous, since no carbon tetrachloride was present. In the preliminary work with acetic acid the bromine addition took place in a solution containing about 45 ml. of water and 15 ml. of acetic acid. Although these solutions appeared homogeneous, erratic results were often obtained. It seems probable that the hexyne was just barely soluble in the solutions and that much hexyne was vaporized in the 300 ml. flask. \/I/hen 30 ml. of aqueous solutions were added ( sodium 2 mm t SAMPLE 3-WAY . '\ -----~ . -f STOPCOCK~ _-Jill II II ~ ~J> 24/40 II II I I Figure 1 Apparatus for Sampling Under Nitrogen w30- Table 6. The Calib~ation of Pipettes for Total Delivery Weight of Mercury in grams Temperature (Deg. ) Volume in milliliters 0.973 13.15, 13.19 20.027, 20.028 26 1.480 1.801 24.42, 24.37 63.12 24( 63.11 27) 79.7, 79.4 21 78.87, 78.95 22 97.7, 95.6, 95.8 24 7.07 157.50 24 11.636 185.43, 185.46 28 13.71 253.8, 253.7 27 18.75 468.7, 468.4 29 34.66 484.4, 485.9 29 35.95 5.88 chloride and potassium iodide) a cloud appeared in the analytical solution. The amount of acetic acid used was in- creased to 30 ml. in the immediately succeeding runs, and gradually increased to 65 ml. during the rest of this work. There seerr.ed to be no rea s on for limiting the amount of acid used for rinsing in the sample. This acid was not the high melting, purified acid but Baker's reagent grade, or, occ~sionally, the forecut from the acetic acid distillations. In addition to the more liberal use of acetic acid rinse, it -31- was made standard procedure to shake the analysis flask intermittently during the five to ten minutes allowed for the brominations. Mr. Dorsey has found that there is no advantage to using sodium chloride but that the potassium iodide should be present in great excess over the mercuric sulfate. chloride was eventually abandoned. The use of sodium The various combinations of reagents used are shown in Table 7. Method I, in _which too little acetic acid rinse was used, has already been referred to. It was used in Runs 1-9. Method II was used for the hexyne determinations through Run 79. It seemed to be entirely satisfactory. Method III introduces the use of 5:1 bromide-bromate solution. _The elimination of the free bromide made it possible to reduce the amount of mercuric sulfate and, therefore, the amount of potassium iodide. Method III was used for Runs 100 to 116 with no trouble, but was found to give poor results when more than two milliequivalents of hexyne were present. The conditions of Method IV were introduced at Mr. Dorsey's suggestion, and with slight modification in Methods VIII and IX were used for Runs 117 to 150. Water was added simultaneously with the potassium iodide because, in more dilute solutions, the starch-iodide endpoint is more readily determined. Moreover, dilution helps prevent the fading endpoint caused by hexenyl acetate. Method V was used for Runs 118 through 126 but was found to give erratic results with higher (>2 milliequivalents) quantities of hexyne. 10-30 VIII c. 20 VII a. b. 20 VI 5 5 5 5 5 5 5 5 5 15 15 4 3 2 3 5 3. 45-65 45-65 45-65 45-65 45-65 45-65 40 30'-35 15 3 15 15 10 15 15 15 15 3 5 5 15 100 100 100 100 100 ¾O 0.1667 Min Br-; 0.01667 Min Bro 0.0833 Min Br-; 0.01667 ~ in Br0 3 , except that some analyses using method III were made with a soTution half this strength. Includes sample and rinse. 5-10 10-20 V IX 20-30 II IV 20.-30 I 10-10 20-30 Method III Br- - Br03 5:lb 10:la Variations in Method of Analysis of 3-Hexyne with Standard Bromide-Bromate Solution. Volumes are in ml. H2S04 HgS04 AcOH 0 HgS04 NaCl . KI 6 M O. 2 M 0.6 lVI 0.6 M 2 M 1 M 1.3 M 3M Table 7. I l\) I vl -33- Table 8. Reproducibility and A,ccuracy of Determinations of 3-Hexyne. Method Hexyne present in aliquot, m. e. Hexyne found, m. e. Accuracy (%) II 1.50 1.58 1.60 1:60 105 107 107 II 1.87 1.97 1.97 105 105 II 2.05 2.18 2.20 2.20 2.19 106 107 107 107 III 2.06- 2.10 102+ III 0.994- 1.026 103-t V 1.970 2.062 2.062 105 105 IV V ?* 1.664 1.668 1.663 1.661 1.650 1.665 1.657 1.653 VIII 2.100 2.220 2.215 2.205 2.213 106 105 105 105 VIII 2.100 2.172 2.187 103 104 V V VI VI VI VII * The amount of hexyne present in these eight determinations was not accurately known. All had the same amount of hexyne, however, and the reproducibility of the "Hexyne found" is considered satisfactory. -34- Methods VI and VII were tried in a special experiment and seemed to be satisfactory. any other time. They were not, however, used at Typical results for most of these methods are shown in Table 8. The calculation of the hexyne present in the aliquots (column 2, Table 8} was made by assuming additivity of volumes when hexyne was dissolved in acetic acid. On this basis it is seen that the hexyne found is about 6% high in these titrations. (The titrations using Method III were made on different dilutions of the same solution of hexyne in acetic acdrl which had been prepared two months before the determinations.} reproducibility, however, is quite good. The The spread in the values of hexyne found from any solution of hexyne is usually less than 1%. One titration of hexyne was made according.to Method II with the mercuric sulfate and sodium chloride eliminated. ,minutes were allowed for the bromination. Ten Only 73% of the hexyne (1.60 instead of 2.20 millequivalents) were found B. Determination of Hexenyl Acetate (3-Acetoxyhexene-3). Method II for the determination of hexyne (3 mil- limoles of mercuric sulfate and 20 millimoles of potassium iodide) was the first method tried for the analysis of hexenyl acetate. An aliquot containing 1.644 milliequivalents of the ester was taken for the determination. When the final iodine solution was titrated with standard thiosulfate solution, it was found that, though the blue starch-iodine color was discharged as expected, it retur~ed very rapidly. _ The thiosulfate burette readin g the first time the solution became colorless corresponded to 1.63 milliequ ivalents of hexenyl acetate a quite satisfactory check. The next day a similar titration was made on the same solution. By workin g rapidly the initial thiosulfate titer was secured about one minute after the addition of the potassium iodide. The indicated quantity of hexenyl acetate was 1.648 milliequivalents. On the third day a similar titr~tion indicated 1.653 milliequivalents of ester. It was concluded that this method of titration gave satisfactory results, and it was used for Runs 48 through 50. For the analyses of Runs 51 through 56, mercuric sulfate was omitted from the analytical solutions. quite satisfactory. The results seemed The method of choice for the analysis of hexenyl acetate is probably to add the solution of the ester to the acidified bromide-bromate solution with enough acetic acid rinse so that the solution is about 75% acetic acid. After five to ten minutes of intermittent shaking, 20 millieqµivalents of potas s ium iodide a n d water to a total of about 100 ml. should be adde4 and the solution titrated rapidly with sodium thiosulfate. Drifting of the endpoint continued to be troublesome, however, and several further variations were tried. A reason- able hypothesis for the continued appearance of iodine in the solutions is shdwn in Equations 17 and 18. -36- AcO H I I AcO H I I Et-C-C-Et 4 HI --~>~ Et-C-C-Et • IBr I J l I H Br BrBr (17) IBr • HI (18) On such a basis the drift should be diminis he d by diluting the solution and by using less potassium iodide, the source of the h y driodic acid. Some analyses were made without mercuric sulfate and with 0.9 milliequivalents of potassium iodide. In these solutions the drift was cut to ne g ligible proportions by adding up to 100 ml. of water with the potassium iodide. However, the results were scattered and low. In one set of determinati·ons, aliquots containing 0.684 milliequivalents of ester gave titers of 0.498, 0.482, 0.530, and 0.515 milli• equivalents. A return to the use of mercuric sulfate and somewhat larger concentrations of potassium iodide was then made. Two combinations of reage n ts were found to give satisfactory results. They are shown in Table 9. Five ml. of sulfuric acid and 60 ml. of acetic acid were present, as usual. Method I was u s ed in t wo deterninations of aliquots supposed to contain 1.50 milliequivale n ts of ester. wer e 1.37 and 1.38 milliequivalents. The quantities found Me t h od II indicated 0.642 milliequivalents in an aliquot supposed to contain 0.685. These determinations were run on hexeny l acetate made with perchloric acid as catalyst (combined cuts C and D, Table 4) and seem to indicate the presence of 6% of saturated impurity. Method II was used in Runs 200 through 208. Table 9. Methods of Analysis of Hexenyl Acetate. are in ml. I 18 II 20* .,i, HgS0 4 KI 0.2 M 0.6 M 1.0 M 3.0 M Volumes 10 5 2 5 100 O. 0500 N C. The Bromination of Diethyl Ketone An analytical solution containing 1.2 milliequiva.- lents of mercuric sulfate (Method II, Table 9) was found to brominate diethyl k etone. To a solution containing 10 ml. of 0.05 N bromide-bromate, 3 ml. (31 milliequivalents) of ketone were added. The bromine was completely reacted, no color being found on the addition of potassium iodide. When the vacuum was broken and the flask exposed to air, a small amount of color developed. An acetic acid s olution 0.50 Min diethyl ketone was then prepared and 1 ml. aliquots ta.ken for 11 determin ... ation" using an equivalent amount (one half a milliequivalent) of bromide-bromate solution. Only about 10% of the bromine was found to react under these conditions. It was concluded, -38- therefore, that small quantities of ethyl propyl ketone, which bears a close resemblance to diethyl ketone and which is produced during kinetic studies, would not seriously affect the analytical data. Some comparable studies were made in which the mercuric sulfate was omitted from the analytical solution. Two ml. (21 milliequivalents) of diethyl ketone reacted with most of 0.5 milliequivalents of bromide-bromate solution, a faint color a ppearing on the addition of potassium iodide. The returning end-point made the titration difficult, but it was found that at least half of the bromine must have reacted with the ketone. Using the 0.50 M solution of diethyl ketone, it was found that no reaction took place between 0.5 milliequivalent of ketone and 0.5 milliequivalent of bromine. Duplicate analyses with two ml. (1 milliequivalent) of ketone solution were also run. One showed no reaction. 'of thebromine reacted. In the other about 3% It was tentatively concluded that under these analytical conditions mercuric sulfate catalyzes the bromination of diethyl ketone. D. The Action of Diethyl Carbitol in the BromideBromate Titration. A-solution containing 0.4 milliequivalents of bromide- bromate solution, 3 milliequivalents of mercuric sulfate, and 20 milliequivalents of potassium iodide (Method I, Table 7) was titrated to the equivalence point with thiosulfate solution. To this solution some diethyl carbitol was added and a large -39- amount of free iodine was immediately released. This test was run one month after the purification of diethyl carbitol described earlier. Even if the indicated oxidizing power resulted from the formation of peroxides within this period and was not an intrinsic property of the pure solvent, it suggested strongly that diethyl carbitol would not be an advantageous choice of S'olvent for this study. -40- IV. The Reaction of 3-Hexyne and 3-Hexenyl Acetate-3 with Acetic Acid. Preliminary Results. A. The Reaction of 3-Hexyne 1. Calculations. When acetic acid adds to hexyne to give hexenyl acetate (Equation 16), the unsaturation as measured by a bromine titration changes from four equivalents to two equiva~ lents per mole of solute. The initial concentration of hexyne could be determined by titration according to Equation 19 if it were possible to make a titration at the moment of mixing the solutions. H0 is the initial concentration of hexyne (19) in moles per liter and U is the observed unsaturation in equivalents per liter. Only a small fraction of the change of unsaturation observed durin g the first few minutes of the reaction need to be ascribed to the subsequent reaction of the hexenyl acetate. This fact is indicated by the isolation of hexenyl acetate from the solution and will be discussed later in connection with studies on the rate of reaction of hexenyl acetate. An approximate material balance can, therefore, be written as in Equation 20. H0 =H• V The moles per liter of hexyne and (20) hexenyl acetate (diethylvinyl acetate) at any time after the start of the reaction are represented by Hand V respectively. The amount of unsaturation in equivalents per liter at any time -41- is dependent on the concentration of both the hexyne and hexenyl acetate as shown by Equation 21. = 4H • 2V U (21) From Equations 20 and 21 V may be eliminated, giving Equation 22. H = U - 2Ho (22) 2 Because the titration of hexyne gives results about 6% high, these equations must be modified. Equation 21 becomes: Ua: 4(1.06)H + 2V (21a) where Ua is now the observed unsaturation. Equation 22 must also be modified: H: Ua - 2 Ho (22a) 2.24 In many runs the value of H0 was determined not by the weight of the hexyne used but from volumetric calculations. For example, in several runs a special pipette WQS used for I adding about half a ml. of hexyne to 99.5 ml. of acetic acid solution. By making several titrations in the absence of catalysts, it was found that the apparent initial hexyne concentration, Hao, was 0.0465 moles/liter , in such a solution. In other runs a special pipette was used for adding 0.27 ml. of hexyne to 20. 0 ml. of acetic acid. Repeat.ed ti trations showed that such a solution had an · apparent hexyne concentration of 0.123 moles/liter of hexyne and that if, for example, five ml. of this solution were added to 95 ml. of reaction solution, the apparent initial concentration would be 0.00615 moles/liter. ... 42- These apparent concentrations are related to the true concentration by Equation 23. H80 = 1.06 H0 (23) Introducing Equation 23 into Equation 22a gives Equation 24. H = Ua - 1.89 Hao (24) 2.24 If the factor of 1.06 is ignored, the concentration of hexyne would be given by Equation 25. H - Ua - 2 Hao (25) 2 The difference between Equations 24 and 25 is often small. 2. Reactions catalyzed by perchloric acid &nd mercuric acetate. Preliminary results. The data collected in ten runs catalyzed by perchloric acid and mercuric acetate are shown in Table 10. These data do not suffice to explain the mechanism of the catalysis, and the various attempts to correlate the runs will not be reproduced here. However, various points are worth com.inent ing on. A comparison of Tables 1, 3, and 10, shows that probably Run 24 or 25, Table 10, repres ent a more promising set of conditions for synthesizing hexenyl acetate than any that wa:s , actually used. Run 359, Table 1 is nearly the same as Run 24 except for the higher concentration of hexyne in Run 359. Evidently the high concentration of hexyne has a detrimental effect on the course of the reaction. In Table 3 lower -43- concentrations of hexyne were used, and Run 19a, which was used for large scale production of hexenyl acetate, actually had less hexyne than Run 24 or 25. It can be seen from Table 10 that the concentrations of catalyst should have been kept some~ what higher than they were in Table 3. Run 73 showed a marked "tapering off" effect. There was a sudden initial drop in the amount of unsaturation followed by a virtually static condition. The same effect was shown, somewhat less dramatically, by Runs 29, 32, 33, and 64. It is, of course, possible that Run 31 also tapered off after a very small initial drop. Run 26 seemed to continue better than the others, but even here the rate diminished faster than would be calculated for a mechanism depending on the zero or first po\~er of the concentration of hexyne. This tapering off effect was noted throughout the work and no satisfactory explanation has yet been evolved for the phenomenon. The runs ln which the hexyne continued to react were those in which the concentration of perchloric acid exceeded the concentration of mercuric acetate and in which the concentration of hexyne was not much greater than O.l !· Runs 21, 27, and 28 are important for showing that neither perchloric acid nor mercuric acetate is, by itself, a very effective catalyst for the addition of acetic acid to hexyne. Yellow (100) 53 (100) 100 100 (100) 96 92 91 89 0 13 0 42 600 0 20 1 5 5 0.5 1 16.2 15.6 16.2 30 25 25.l 30 31 32 78 15 29 3 Gr. Yellow (100) 87 74 72 0 2 17 40 11 2.5 15.8 30 29 (100) 64 45 16 0 4 28 4300 22 5 15.8 30 26 400 2 37 178 81 (100) 21 0 2 145 11 days 60 Gr. Yellow Slight Colorless Yellow Rd. Brown Yellow Lt. Brown Yellow Color 10 15.8 30 25 400 182 163 152 120 Unsatura.tion Equivalents/ liter x 103 (100) M x 103 Hexyne 0 1 day 2 days 3 days 7 days 30 24 HClOi M x o3 Time Min. Catalysis by Me rcuric Acetate 40 (Deg.) HOAc Hg(OAcJ2 M x 10 M. P. (Deg. ) 10 16.0 Run No. Temp. The Rate of Reaction of Hexyne and Acetic Acid. and Perchloric Acid. Table 10. I I ~ ~ 1 1 16.6 Excess AC20 Excess AC20 16.0 25.1 25 25 30 30 30 33 64 73 21 27 28 (42.8) 31.6 29.6 30.0 (100) 100 99.5 98. 7 (100) 100 101 {120) 100 100 98 0 6 30 62 0 13 100 1300 0 5 160 0 6 0 2 20 3 4 0 0 15.8 16.2 90 1300 4 (42.8) 39.4 37.6 36.7 0 9 47 106 6 Hexyne M x 10 3 (100) 94 92 91 90 89 Time Min. 0 4 16 31 59 134 HCl04 M x 103 4 (De g .) - M.P. (Deg. ) Hg(OAc)z M x 103 No. HOAc Temp. Run Table 10 (Continued) Unsaturation Equivalents/ liter x 103 None Colorless Gr. Yellow Color I I t,f::,. (JI -46- 3. Reactions catalyzed by boron trifluoride and mercuric acetate. Because boron trifluoride had proved to be such a successful cata,l yst for the synthesis of hexenyl acetate, a number of kinetic studies using boron trifluoride and mercuric acetate were made. They were notably unsuccessful. In several cases runs had to be abandoned or results discarded because precipitates were formed from the solutions. Of the "success- ful" runs the highly erratic nature can be seen from the brief summary in Table 11. Boron trifluoride is an unusual acid and, in some ways, an exceptionally strong one. Ether and mineral acids are known to form loose complexes which do not withstand distillation. The ether-boron trifluoride complex, on the other hand, was purified by distillation. It is probable that slight amounts of water or acetic anhydride in the reaction solutions had Ii profound effect on the availability of the boron trifluoride. Moreover, the presence of the ether, added with the acid, was an additional complicating factor. The attempt to study the reaction of hexyne with acetic acid using boron trifluoride as the acid catalyst was abandoned. 4. Reactions catalyzed by sulfuric acid and mercuric acetate. The data from nine runs catalyzed by sulfuric acid and mercuric acetate are shown in Table 11. Two ways in which these runs differ from the runs with perchloric acid -47- Table 11. The Rate of Reaction of Hexyne and Acetic Acid Catalysis by Mercuric Acetate and Boron Trifluoride Run No. Hg(0AcJ2 M X 10 BF3 M x 103 Hexyne M x 103 Hexyne reacted in one hour(%) 41b 4.54 9.09 43.6 8 42a 5 20 50 9 42d 5 20 50 49 44a 3 20 50 75 45b,c,d 1 20 50 5 seemed to be significant. First, no colors were associated with the sulfuric acid runs. Second, and perhaps not coincidentally, the runs with sulfuric acid did not show the tapering-off phenomenon to anywhere near the extent that the runs with perchloric acid did. The rates did taper off, however, faster than could be accounted for on the basis of zero or first order ,dependence of the rate on the concentration of hexyne. The absence of the by-product and the less marked divergence of the rates from zero or first order dependence on the hexyne, made it seem possible that the runs with sulfuric acid would be more amenable to interpretation than the runs with perchloric acid. It was known tht the catalyst would be complexed by the initial product, hexenyl acetate, and that a tapering off in the rate of reaction of hexyne was to be expected. It was assu_med that equilibrium reactions took place between the mercuric acetate and both hexyne and hexenyl acetate. No information was available at the time these runs 1 Exce s s Ac20 250 66 1 Excess Ac 2 o 25° 65 5 16.0 30° 20 - Hg(OAcJ2 M x 10 5 30° 19 HOAc M.P. (De g . ) 16.0 Temp. No. Run - 10.3 7.35 1 1 ~soi 0 3 Mx 0 1 28 48 69 161 0 6 48 78 137 0 17 56 120 182 1250 2650 5 dliys 15 50 940 1060 2340 2750 0 Time Mi n . - (45.0) 42.4 37.0 33.9 31.2 27.2 {44.5) 43.2 35.7 32.4 30.7 81 74 88 (100) 92 (100} 91 82 66 Hexyne M X 103 The Ra te of Reaction of Hexyne and Acetic Acid. Catalysis b y Mercuric Acetate and Sulfuric Acid. Table 12 210 142 16 (400) 230 150 130 (400) Unsatµrati0h Eq. / 1 • x 103 I ~ I Cl) Temp. 25° 25° 25° 25° 25° 25° Run No. 67 69 70 71 72 13 16.4 Excess Ac20 Excess Ac 2 o 0 1.39 1.39 (0) 0.78 160 200 5 0 6 24 45 0 6 30 53 5.88 4.41 42.0 36.5 32.7 30.0 7 27 47 66 (44.5) 0 11.76 1 (44.5) 42.7 39.0 36.5 (78) 77 77 76 (44.5) 43.0 38.8 36.7 41.9 38.0 34.9 32.9 Excess Ac20 (44.5) 42.6 39.4 37.1 35.6 (44.5) Hexyne M x 103 0 7 26 53 77 ( 0} . 12 37 67 103 Time Min. 8.82 4.4 ivi2~fo3 1 1 Hg(OAc)2 M x 103 Excess Ac20 Excess Ac 2 o fuP• e g .) HOAc - Table 12 (Continued) Unsaturation Eq./ 1. X 103 I c.o I ~ -50- were made as to the exact composition of the complexes. Hugel 5 and Hibou (See Section I) had written Cn¾n"(Hg(OAc) 2 ) for the complexes of simple alkenes in acetic acid. In analagous fashion, the complexes for hexyne and hexenyl acetate were now written c6H10 •Hg(OAc) 2 and C6H110Ac Hg(OAc) 2 • 0 are shown in Equations 26 and 27. The equilibrium const~nts Their ratio is to be denoted by n as shown in Equation 28. (26) (Cv) = (Cr)(V) (27) n = Kv KH (28) CH, Cv, and Cr are, respectively, the concentrations of the hexyne-mercuric acetate complex, of the hexenyl acetate1mercuric acetate complex, and of the free mercuric acetate. Because the hexyne is used in large excess over the mercuric acetate and because of the known stability of the mercuric complexes, Cv• Cr is believed to be small compared to CH and to This approximation is expressed in Equation 29, in which C0 is the total mercuric acetate. The hypothesis was then made that the rate of disappearance of hexyne is proportional to the mercuric complex, as shown in Equation 30. (29) dH -- k CH ... at (30) Combining Equations 26 throu gh 30 we get Equation 31. dH • kC 0 - at (31) -.,.----v n H• 1 V and H may now be replaced by U and H0 by means of Equations 20 and 21. 2KC 0 dU dt = n( u - 4H0 u - 2H0 (32) ) + l Integration and simplification give Equati on 33 in which 0 F(u) = n Ho Cn U -2H2Ho - ( n- l) 4Ho - u 2 t = 1c K o (33) Uhas been i.ntegra.ted between 4H0 and U and t from Oto t. Equat·ion 33 was applied to runs 65-72 with various values 1 of n being inserted. The left hand side of the equation was then plotted against t and the curves examined to see which value of n gave the best straight line. The curves for runs 69, which are typical, are shown in Figure 2. If only the origin and the first three points be considered, n: 9 gives a convex curve, n = 7 gives a strai ght line, and n = 5 gives a concave curve. In all cases the fourth experimental point devi ates markedly. I n creasin g the value of n to bring this point into a straight line would make the first part of the curve increasingly convex. Actually the curve for n = 1 -52- seems to be the smoothest curve and would indicate that some factor not being considered was responsible at least in part for the tapering off. It was found that the value of the first point in the plot of F(U) against time was rather insensitive to the value of n selected. This fact su ggests an alternative method of finding the value of k, Equation 30. If the value of n is, somewhat arbitrarily, set equal to one, then Equation 33 reduces to Equation 34. Accurate determinations should be made of H0 and of some value of Hat a value oft low enough so that the branches of Figure 1 have not yet diverged. The rate constant could then be evaluated from Equation 34. The data of Ta ble 12 are not suitable for this purpose because the values of H0 were not known with sufficient accuracy in these •runs. In addition to the unknown factor causing the tapering off, at least two other factors were not incorporated into the above equations. It seemed probable 7 that the acid catalyst, known to be necessary for the occurrence of the reaction at a significant rate, would play a role in the formation of the complexes. matter. Equations 26 and 27 take no cognizance of this Since neither KH nor Kv but only their ratio appear in the subsequent equations, if acid happened to appear in the same way in the two equilibrium expressions it would cancel -53- P{U) 0.028 / n = 9 / n .. 7 0.024 n = 5 0.020 n = 3 0.016 I 0.012 / 1/ n = 1 Fi gure 2 Run 69 0.008 F( U ) Plotted against f; 0.004 Time, in Minutes. See Equation 33. 0 0 20 40 60 80 100 12 0 Ti ,,.e in minu te s .. 54_ out in Equation 28. In this case the later equations would still be correct but data on this point were not available. Allied to the question of the part played by the acid in form... ing the complexes was the question of how much free acid remained in solution. It seemed conceivable that this free acid played a role in the addition reaction. Since the hypothe- sis proposed in Equation 30 was not established by runs 65-72, an altogether different mechanism, depending on the free acid in solution, remained an active possibility. Some intimate method of determing the "location" of the available acid was therefore, needed an~ it was for this reason that the acidimetric work, discussed in Part V, was undertaken. 5. Studies relating to the yellow color appearing in solutions of hexyne, mercuric acetate, and perchloric acid. Acetic acid solutions of hexyne, mercuric acetate, and perchloric acid were nearly always observed to be colored. Investigation of this color were made first because it was recognized that the color might be an intermediate perhaps a mercuric carbonium complex - in the production of hexenyl acetate. This possibility seems to have been eliminated. Interest in the yellow material remains, however, both because it is an important by-product of the reaction being studied and because the material is of interest in its own right. Neither the isolation nor the concentration of the product have been -55- accomplished and its structure is unknown. Information has been obtained, however, on the mechanism of its formation and its rate of appearance and disappearance. Nine hours after the reaction had started, the solution of Run 30, Table 10, was put in a 1 cm. corex cell and the absorption curve of the yellow by-product determined by means of a Beckman spectrophotometer. Absorption peaks at 360 mf (D = 1.52) and at 352 mf (D = 1.512) were observed. The peak at 352 mf was used for following runs Cl to C6. In a set of preliminary runs it was found that the color density increased to a maximum in about 25 minutes and then decreased at a quite constant rate. Runs Cl to C6 were made to get preliminary data on this rate of disappearance and readings were not made on most of the solutions until about 25 minutes after mixing, when the peak of color density had just been passed. The data are presented principally as a guide to further work if this research is continued. In addition to the runs shown in Table 13, three determinations were made which are not shown in which the maximum density was very much less than expected from comparable runs. It is believed that water may have contaminated the solutions in these cases, and it was found that water acted rapidly to destroy the yellow color. 0.5 0.5 0.5 0.5 0.3 200 150 200 300 300 C3a C4a C3c C5a C6d B A 0.75 1 Hg(OAc~2 100 100 M x 10 Hexyne M x 103 Cl C2 Run No. 2 2 1.2 2 2 3 4 C HC104 M x 103 0.698 0.634 0.213 0.701 0.740 1.35 1.83 D Maximum Density 0.0038 0.0034 0.0030 0.0034 0.0028 0.0027 0.0025 0.0030 0.0021 0.0006 0.0028 0.0037 F E 0.003s 0.0067 E D in Density dimunution per min. Observations of Yellow By-product at 352 mf Anhydrous Acetic Acid. Table 13 0.70 0.63 0.59 0.70 0.74 0.60 0.46 G D B X C I ' (J) en -57- The runs in Table 13 were made in acetic acid of M.P. 16.6° prepared with acetic anhydride. Excess acetic anhydride was used in Ru..ns C3c, C5a, and C6a. Since no temperature control was attempted, the runs which were made at the same time are shown grouped together. Column F, Table 13, was an attempt to correlate the rates of disappearance of the color in different runs. Since it seemed reasonable that the rate of disappearance of the. color would be proportional to the color, the rate (Column E) was divided by the maximum color density (Column D). The quotient, shown in Column F, is reasonably consta r:.t for the different runs, and the insertion of other figures from the table into this calculation does not improve the agreement. It may be, therefore, that the yellow compound breaks down spontaneously or reacts with the solvent without catalysis. In either case the difficulty of isolating the material would be predictable. Column G is an attempt to use the data to explain the formation of the color. If the amount of color formed depended on the concentration of the mercuric acetate and the perchloric and but not of the hexyne, the ratio in Column G would be a constant for the different runs. except for Run C2. The a greement is fairly good That the maximum density for Run C2 is slightly low shows up in both Column F and Column G. The discrepancy may be due to a trace of water or it may be the result of the high rate of color disappearance in this run. -58- The high rate, operating during the 25 minutes the peak color was being built up, would, of course, lower the maximum value obtained. The runs shown in Table 13 were all made with a considerable excess of perchloric acid over the mercuric acetate. A set of runs in which this condition did not obtain are shown in Table 14. These runs were made in acetic acid rectified by distillation, m.p. 16.6°. 410 mf. The observations were made at It should be noted that Runs C9 and ClO, also Cll and Cl4 are duplicates and that the reproducibility is fairly good. All runs were made with 0.0305 M hexyne. ,At a constant concentration of mercuric acetate (Runs C7 to Cl2) the maximum optical density attained is proportional to the concentration of the perchloric acid, though the proportionality constant drifts when the mercuric acetate is in considerable excess over the acid (Column G). Surprisingly, 'however, the time needed to attain this maximum optical density is a constant independent of the acid concentration. Column H is a not too successful attempt to relate the rate of color dimunution to the maximum color. This rate seems to diminish with the acid concentration. Four Runs (Cl3 to Cl6) were made in which the concentration of perchloric acid was constant and that of mercuric acetate varied. was In all but the first of these runs the mercuric acetate in excess. Strange results continued to be the rule, for it was found that with increasing concentration of mercuric Run 20 0.076 1.21 6.90 30.5 Cl6 23 0.146 1.21 4.14 30.5 Cl5 25 ... 35 0.241 1.21 2.07 30.5 Cl4 45 0.283 1.21 0.69 30.5 Cl3 0.0010 0.17 .00015 27 0.150 0.86 2.07 30.5 Cl2 0.0012 0.20 .00029 25-35 0.239 1.21 2.07 30.5 Cll 0.0017 0.21 .00060 26 0.362 1.72 2.07 30.5 Cl0 0.0015 30.5 C9 0.22 1.72 2.07 30.5 C8 .00055 21 0.370 2.24 2.07 30.5 0.0016 H F D 0.22 .00080 22 0.485 G D C 0.0018 .00100 23 0.562 2.58 2.07 F Time at Density dimunMax.D.,Min. ution per Mi n . E D Maximu.'11 Density C HCl.Oi M x 03 B Hg( OAci 2 M x 10 0.22 A Hexyne M x 10 3 C7 No. Observations of Yellow By-product at 410 mf . Table 14. I c.o l CJl 0.072 0.390 0.066 0.356 0.220 0.75 1.50 1.00 1.00 0.31 1.50 1.00 1.00 0.50 0.33 5.8 5.8 11.6 5.8 3.5 375 360 360 360 360 360 360 Cl30 Cl31 Cl32 Cl33 Cl35 Cl36 Cl37 Cl38 43 35 50 35 30 44 0.466 0.470 0.215 0.461 0.210 0.519 0.239 0.238 0.105 0.270 0.112 0.242 o.oo 0.221 (0.100) 0.247 0.106 0.226 o.oo 0.50 1.00 0.50 1.50 0.50 1.00 0.50 0.50 1.50 3.5 5.8 5.8 360 360 Color Time 40 0.072 Maximum Density o. 681 C21 0.035 0.032 0.47 0.67 0.21 375 C20 0.061 0.056 0.60 0.18 0.64 375 Cl9 0.081 0.50 0.53 375 Cl8 0.80 0.42 375 Cl7 0.074 Color 2 min. 2½ min. 0.50 HCl04 M x 103 0.059 - Hg(OAcJ2 M x 10 0.056 Mf • Hexyne M x 103 0.40 Run No. Observati ons of Yellow By-product in Acetic Acid M.P. 16.6° Table 15 I 0 I CT) --61- acetate, the maximum absorption peak was reached more quickly, but that the maximum optical density obtained was lower. Table 15 shows two sets of later data. Runs Cl7 to 021 were made in connection with indicator work to be described later. The conditions were deliberately random and no con- clusions seem to be derivable from them. here to make this section complete. They are included Runs 0130 to Cl38 were made in connection with kinetic work to be discussed later. These runs show a very neat pattern. In six runs (0130 to 0136) the concentration of mercuric acetate equaled the concentration of perchloric acid. The amount of color developed at both 2 1/2 minutes and at maximum optical density is proportional to this single concentration. in Table 16. The relation is shown The depende nce of the amount of color on the amount of catalyst and its independence of the excess of hexyne present suggests strongly that a complex of hexyne and the two catalysts 'is the intermediate in the formation of the by-product. Cl~7 and Cl38 diverge from this pattern. Runs Run Cl37, with excess pe.r chloric : acid, gave more color than would be expected . if the concentration of a 1 to 1 to 1 complex were the only concentr2tion of significance. Run Cl38, with three times as much mercuric acetate as perchloric acid, had no color at all. A complete investigation of the absorp:tion spectrum of the yellow color was made with the solution from Run 130. Because the color density was not constant, a special technique was employed. Two and a half hours after the solution was mixed, -62- . Table 16 Mechanism of Formation of Yellow By-product No. Concentration of Catalysts Hg(OAc)2: HCl04 Cl30 Cl31 Cl32 Cl33 0135 Cl36 0.0015 0.0010 0.0010 0.0005 0.0010 0.0005 Run Maximum Color at 360 mt'-. Rstio of Color to Catalyst Concentration o. 681 450 470 470 430 460 420 0.466 0.470 0.215 0.461 0.210 a rapid set of readin g s were made at 10 m.,«, 400 mt' to 320 m,-,.. • line. intervals from These readin g s were considered a base The optical density of separate portions of the spectrum were then determined at 1 mp. intervals. Such detailed readings always included one of the base line readin gs. The determina- ti0ns could then be related bae~ to the standard time, 2 1/2 hours, by simple proportion. This procedure assumes that the optical densities at diff er e nt wave lengths maintain a constant ratio with each other as the color fades. To justify this assumption a new set of readin gs of the optical density of the solution was taken the next day. ratios are shown in Table 17. The reading s and their It is evident that at wave lengths longer than 350 m}"- the assumption is justified. The color behaves as though derived from a single moleculsrspecies. The curve at 2 1/2 hcurs is shown in Figure 2. From the data the ratio of the optical density at 360 mr,- to that at 352 mf, 375 mp-, and 410 mf'- is respectively 1.013, 1.200, and 2.663~ Tables 13, 14, and 15 may thus be related. Table 17. The change of Optical Density of the Yellow By-product with Time at several Wave Lengths Wave Length m I'- • 440 430 420 410 400 390 380 370 360 350 340 330 320 Optical Density 2 1/2 hrs. 28 hrs. (0.044) (0.077) (0.137} (0.210) 0.285 0.371 0.449 0.520 0.560 0.525 0.465 0.465 o.604 0.028 0.050 0.088 0.133 0.184 0.238 0.288 0.333 0.359 0.342 0.314 0.331 Ratio 1.57 1.54 1.56 1. 58 1.55 1.56 1.56 1.56 1.56 1.53 1.47 1.40 To further demonstrate that only one molecular species ~~sinvolved,three different peaks (at 345 mf and 387 ml"- , 360 mf , ) were followed during the appearance and disap- pearance of the color. Run 0132 was repeated and the optical densities at these three points followed over a three hour I period. The results are shown in Table 18. Because the ratio between different parts of the curve remained constant as the color developed and faded, it was concluded that only one species was contributing the color to the solutioni. The effect of water and acetic anhydride on the develop~ ment of the color were studied in Runs 141 and 142. These runs resembled Run 0131, Table 15, except that Run 141 was 0.277 Min water and Run 142 was 0.265 Min acetic anhydride. No color readings were made during the first hour or these runs, since this time was taken up with bromate titrations. 0 0 . 100 0 . 200 0 . 300 0 " 400 0.500 0 . 600 340 - Fi gur e 3 '-,/ 350 360 \ 380 ~ I\ 370 /\J \ /' Ex tin c tion o f So l u t ion c f Run 130 2½ Ho ur s a f te r Mixing. V .._, ~/ Op t i cal De n sity ~ - \ 390 \ 400 ' '\. - "' '-- '-- " 410 42 0 wave len g t h in mr< - "" " ' I .:,.. I 0) -6ffi- Table 18. The Ratio of Optical Densities of Three Absorption Peaks of the Yellow By-product during the Growth and Disappearance of the Color in Run 0132. Time Min. Optical Densities 360 345 mf 387 mf mr Ratio of Optical Densities 5 0.328 0.310 0.240 1.37:1.28:l 10 0.400 0.372 0.291 1.37:1.27:1 40 0.473 0.440 0.346 1.37:1.28:1 189 0.433 0.403 0.316 1.37:1.29:1 After 93 minutes, however, Run 141 had an optical density of only 0.165 at 360 mf. The density rose slowly to 0.197 at the end of four hours. (Run Cl3l reached a maximum optical density of 0.466 in 43 minutes.) Run 142, on the other hand, showed virtually no deviation from Cl31. Readings of the optical density at 360 mf after 115 and 130 minutes were 0.442 and 0.442. The absorption curves of solutions of each of these I runs showed the characteristic 1.37:1.28:1 ratio shown in Table 18. The rate of change of unsaturation in these runs differed very little from that observed in Run 131. In Table 14 there was noted a tendency for the rate of dimunution of color to be less when the acid concentration was low. As part of the general attempt to isolate the yellow product, some potassium acetate was added to the solution of Run Cl30 to see if, in the absence of acid, the color would be stable. Two and a half hours after the start of the reaction and an hour and a quarter after the color peak of 0.681 had -66- been reached (see Table 15), enough potassium acetate, previously dried at 110°, was added to some of the solution to give a 0.1 M solution of the salt. then determined. The rate of color diminution was The data are shown in Table 19. It is evident that, despite the data of Table 14, potassium acetate increased Table 19. The Effect of Potassium Acetate on the Rate of Disappearance of the By-product. Run 0130. Time Min. Optical Density of Solution 360 mr, 345 ' 387m 40 70 145 160 177 182 0.681 0.664 mr 0.620 0.611 0.609 0.580 0.451 0.570 0.442 Optical Density of KOAc Solution 360 mr, 345 m f 387 mf' KOAc added 0.598 0.556 0.555 0.541 0.509 the rate of disappearance of the yellow material. 0.431 0.394 The curves of optical density against time have a round top, and it is possible that the rates of diminution shown in Table 14 represent ' merely the portion of the curve between the peak and the portion of linear decline. Alternatively, it is possible that the potassium acetate, which deliquesces rapidly, was wet. Water destroys the yellow color rapidl:y-. In an attempt to isolate the yellow factor 15 ml. of an acetic acid solution 0.44 Min hexyne, 0.0024 Min perchloric acid and 0.00065 Min mercuric acetate was prepared. After 20 minutes ligroin (B.P. 30°-60°) and aqueous sodi~~ hydroxide were added. Most of the color was lost when the alkali was -67- added, but the ligroin layer was separated and found to have a very faint yellow color. As shown in Table 20, enough of the absorption curve was taken to show tha t it was different from that shown in Figure 2. Even if the product were unchanged, of course, the change in solvent might have such an effect. Table 20. Absorption Spectrum of Yellow Color Extracted by Ligroin Wave Length (mt) Optical Density 0.066 0.096 0.150 0.160 0.146 410 400 380 370 360 Twenty-two ml. of acetic acid solution were prepared 0.040 Min hexyne, 0.0091 Min mercuric acetate, and 0.023 M in perchloric acid. After 20 minutes carbon tetrachloride and water were added. Color was present in both phases. The phases were separated and the aqueous phase again extracted with carbon tetrachloride. the phases. Again the color was divided between After separation more carbon tetrachloride and sodium hydroxide were added to the aqueous phase. disappeared. The color The first two carbon tetrachloride layers were combined and split. Part was shaken with aqueous sodium bi- carbonate and part with water. In each case the color disappeared. A solution 0.096 Min hexyne, 0.025 Min mercuric acetate, and 0.050 Min perchloric acid was prepared. After 20 minutes -68- 20 ml. of n-heptane and 30 g. of solid sodium bi carbonate were added. The bicarbonate dissolved quite slowly and the solution was warmed slightly. The mixture finally turned solid. Separa- tion of the phases was attempted by centrifugingi which was not successful, and by filtration with suction, which yielded about 3 ml. of slightly yellow solutioo. The absorption curve of this solution wa:s featureless, dropping from an optical density of 0.76 at 336 mf to 0.057 1at 386 mf- and tapering off to O. 037 at 430 m}'--. A similar solution (0.096 Min hexyne, etc.) was prepared and a few tenths of a gram of potassium acetate added after 20 minutes. The solution was then passed through a chroma- tographic column packed with activated alumina. No color appeared on the column, and the effluent solution was also colorless. Only a limited number of formulae for the yellow material seem to be available even for conjecture. The yellow material is probably not an intermediate in the formation of hexenyl aceta t e, for the color reaches a maximum intensity in 20-30 minutes, after the rate of change of unsaturation in the solution has tapered off markedly. due to a polymer of hexyne. The color is probably not The rate of color formation and the amount of color formed ar:e independent of the excess of hexyne present. Moreover the amount of color formed is nearly proportional to the first power, and certainly not proportional to a higher power, of a su ggested 1:1:1 corrplex of hexyne, ... 69- mercuric acetate, and perchloric acid. Moreover, the extreme lability of the color to alkali and even to water su ggest that a hydrocarbon is not the colored species. None of the known products of the reaction is responsible in an simple way for the color. Hexenyl acetate, diethyl ketone, acetic anhydride, and even propylene diacetate do not give a color with the catalysts in acetic acid. The amount of color formed is proportional to the hypothetical 1:1:1 complex as indicated in Table 16. This fact suggests strongly that a primary reaction product is either itself the colored species or is con verted to the yellow product by a fast reaction. One species which seems to fit this speci- fication is hexenyl perchlorate, Et-CH=C(Cl04)-Et. This compound might be formed by the addition, to the 1:1:1 complex, of excess perchloric acid or of the acid in equilibrium with the complex. The formation of this compound would explain many of the observa•tions, such as the tapering off phenomenon, the fact that excess acid reduces the tapering off, the dependence of the amount of color on the excess of perchloric acid, the fact that excess mercuric acetate inhibits the color (though why the color should drop to zero in Cl38 is not clear), and the dif f iculty in isolating the yellow product. However, Run 138, with no color, also showed the tapering off phenomenon, and it is not evident why hexenyl perchlorate should be colored. Organic perchlorates are. not unknown 29 but no vinyl perchlorates seem to have been reported in the literature. -70- B. The Reaction of 3-Acetoxyhexene~3 with Acetic Acid Catalyzed by Mercuric Acetate and Boron Trifluoride. The hexenyl acetate synthesized with mercuric acetate and boron trifluoride catalysts was used in a set of kinetic determinations with these same catalysts. The reaction (Equation 17) was followed by measuring the decrease in the amount of unsaturation. As with the hexyne runs with boron trifluoride, trouble was encountered with precipitates and erratic results. It is evident (Table 21) that Runs 52 and 54 proceeded many times more rapidly than the other runs, a fact that is difficult to interpret on the basis of the small increase in mercuric acetate. If the rate is indeed that sensitive to the concentration of mercuric acetate, then it becomes impossible to explain why Runs 52 and 54 give identical results. The data for these two runs fit very closely the first order Equation 35, where t i s in minutes and Vis hexenyl acetate. For the same re~sons that were discu ss ed in connection dV: 0.0073V ~ (35) with hexyne, namely, the probable sensitivit y of the activity of the boron trifluoride to the ether as well as to possible impurities, this phase of the investigation was abandoned. -71- Table 21. The Rate of Reaction of Hexenyl Acetate with Acetic Acid at 250 in the Presence of Excess Acetic Anhydride. Catalysis by Mercuric Acetate and Boron Trifluoride. No. Hg(0AC)2 M x 103 BF3• Et20 M x 103 Time Min. 48 1 20· 49 1 20 50 1 20 51 2 20 .7 29 58 156 4800 10 200 7 36 83 148 181 8 29 Run 94 52 3 20 54 3.5 20 144 202 368 9 32 48 70 90 9 24 41 58 81 (cloudy) 55 56 none none 20 6 20 22 40 78 9 31 120 1500 2900 Hexenyl Acetate M x 103 75.8 75.3 75.5 75.4 52.0 31.4 30.7 17.7 17.6 17.5 17.4 17.3 18.3 18.4 18.1 17.9 17.8 17.3 17.4 14.3 12.8 11.0 9.5 17.2 15.4 13.6 12.0 10.5 17.5 17.6 17. 5 17.5 17.7 17.6 17.7 16.9 16.8 -72- V. The Measurement of Basicity in Glacial Acetic Acid. A. Preliminary Investigations with Sulfuric Acid. The study of the rate of addition of acetic acid to hexyne, discussed in Section IV, led directly to the acidimetric measurements to be described below. Knowledge of the amount of free acid in the kinetic solutions was needed both because of the role this acid might itself play in the rate of addition, and because, by difference from the known amount of acid added, the amount of acid entering into the various complexes might be determined. Because perchloric acid gave a yellow by-product in the kinetic studies and because runs catalyzed with perchloric acid tapered off sharply for unknown reasons, it was planned that subsequent kin e tic studies would be made with sulfuric acid. Acidimetric studies were, therefore, undertaken with sulfuric acid. It was decided to use indicators for this investigation. Hammett 30 and various collaborators and Hall and Spengeman 31 had investigated the use of indicators for acidimetric work in formic and acetic acids. It had been found that substituted nitroanilines (a list is given in Physical Organic Chemistry 32 p. 266) were suitable indicators. They enter a rapid, reversi- ble reaction with the solvated proton, forming aniliniurn ions, which are colorless. Materials and apparatus. The indicator chosen for this initial work was 2-nitro-4-chloroaniline. The acetic acid and sulfuric acid solutions used had excess acetic anhydride in them. -73- After a few days spent in collecting erratic data, the conclusion was reached that the acetic ahhydride was reacting with the amine. A fresh batch of acetic acid was prepared in which the acetic anhydride used was less than the amount calculated to react with the water present. From some of this solvent a sulfuric acid solution w2s prepared with 96% sulfuric acid. Eastman chloronitroaniline was dissolved in boiling water, filtered hot, and two crops of crystals collected as the solu... tion cooled. at 116.0°. The two crops melted indentically and very sharply Some of the early readings of indicator density (Tables 22 and 23) were made in B. Klett Colorimeter., which does not permit selection of wave lengths and the results of which do not seem to be highly reproducible. A change was soon made to a Beckman Spectrophotometer. One te nth of a gram of the amine was dissolved in 25 ml. of acetic acid. This solution was labeled A. One tenth ml.of 'A was added to 50 ml. of acetic acid (solution Al) and the absorption spectrum investigated in the Beckman Spectrophotometer. A maximum wa.s found at about 410 m f' at V'Thich point the optical density of this solution was 0.201. The top of the maximum was quite flat - readings between 406 and 412 mr,- being indisting- uishable - a fact which makes the readings highly reproducible and well suited for indicator work. The solutions were also quite stable, Al showing no change at all in an hour. Investigation of weak bases. Hexyne was the first material to be investigated for its effect on the acidity of the solutions. -74- A 0.10 M solution of hexyne was prepared and mixed with various amounts of 0.10 M sulfuric acid as shown in Table 22. Indicator Cl was prepared from 2 ml. of A and 50 ml. of anhydrous 1..cetic acid. If the hexyne reacted with any appreciable portion of the acid, the optical density of the solution would depend on the concentration of the hexyne. It is evident that no such effect is operating in these solutions. Hexenyl acetate (from the boron trifluoride synthesis) and di ethyl ketone were also tested for their effect on the acid-indicator system. The results of a test with hexenyl acetate are shown in Table 23. Because the optical density of the solution changed so rapidly that a satisfactory initial reading could not be obtained, the solutions were kept overnight and the readings obtained after this period of standing a.re also shown. It seemed necessary to assQ~e that a reaction WQS taking place, and a reasonable assumption was that, in the I presence of sulfuric acid, acetic acid was adding to hexenyl acetate to give hexanone and acetic anhydride as shown in Equation 17. One or both of these products could then react with the indicator, the acetic anhydride to form an anilirle, the ketone to form a 5chiff 1 s base. By eliminating some of the indicator from the aniline-anilinium ion equilibrium, either of these reactions would raise , the optical density of the . solytion. If this interpretation of the change of optic~l density is correct, it demonstrates that the , hexanone diacet~te is indeed unstable in this system. The test with diethyl ketone is shown in Table 24 and it is evident that here too a reaction took place. Table 22. The Effect of Hexyne on the Acidity of Solutions Containing 4-Chloro ... 2-nitroaniline. Hexyne 0.10 M H2S04 0.10 M 15 10 5 10 ml. Indicator A ml. Cl 5 5 5 5 5 10 15 0.2 0.2 0.2 0.2 5 10 HOAc Optical Density (Klett) 0.253 0.265 0.264 0.260 5 5 5 0.248 0.251 Table 23. The Effect of Hexenyl Acetate on the Acid-IndicRtor System. Volumes are in ml. H2S0 4 0.10 M HOAc Cl 5 5 10 5 5 10 Optical Density (Klett} 5 min. 24 hrs. 0.242,4 0.254,9, 0.261 0.252 0.325 Table 24. The Effect of Diethyl Ketone on the Acid-IndicatoT System. Volumes &re in ml. HOAc H2S0 4 0.10 Cl Optical Density, 410 m~ 5 min. 5 hrs. 24 hrs. 0 5 5 5 3 3 0.504 0.407 0.540 o. 588 It is possible that a Schiff's base was formed by a reaction between diethyl ketone and chloronitroaniline. B. Extended Investigations with Perchloric Acid. 1. Introduction and theory of calculation. It was decided to make an extended investigation of the indicatcr method of determining the stren gth of weak, colorless .bases in glacia 1 acetic acid •. The method had been introduced b y Hammett and Deyrup 33 who made a single measurement on propionitrile and a single measurement on water. They used b.enzene sulfonic acid as their stron g acid and f o rmic acid as the solvent. No further use had been made of the method• which seemed to. be suitable for mea s uring the basic strength of bases too weak or too insoluble to be measured readily in water. Hammett and Deyrup found that sulfuric acid was a completely ionized monobasic acid in formic acid and that benzene sulfonic ' acid was nearly as strong. In acetic acid, however, Hall and Spengeman found that the mean activity coefficient of sulfuric acid is strongly dependent on the concentration. Perchloric acid is known to be stron ger than sulfuric acid, and it was used in the present work to minimize the effect of c oncentration on acidity. A·solution consisting of solvent and an acid s olute has a certain tendency to give a proton to a neutral base in solution. This tendency may be called the acidity. The -77- ability of a neutral base to take a proton from the s olution is measured by the ratio (BW)/(B) , and this ability is proportional both to the acidity of the solution and the strength of the base, Kb. The symbols Band BH* refer to the neutral base a nd its conjugate acid. Parentheses, (), signify con- centrations, and brackets, ( J, will be used for activities. Basic strength, Kb, is defined b y Equation 36. Kb : 1 Ka (36) = If the proportionality between the ratio (BH~)/(B) on the one hand and the basic strength of the base and the acidity of the solution on the other is set equal to unity, we may write Equation 37. (BH 4 ) {37) (B) The symbol his a measure of the acidity of the solution. Ta.king the logarithms of Equation 37, and defining Hp= -log h, we get Equation 38. Ho: log ~ + pK 2 (38) Here Ho is Hammett's acidity function 32 • There are both theoretical and experimental reasons for believing that the acidic tendency of the solution may be represented fairly accurately b y Equations 37 and 38. When a base is also an -78- indicator, we will write I and IH• for Band BK'. If an indicator and another base are present in the same solution, the acidity will, of course, be the same for each base. Eliminating Ho from Equation 38 we may write Equation 39 for such a solution. log (I) (39) (IH-1) It is necessary to distinguish between Kb and K8 , the thermodynamic, or activity, constants of Equation 36, and the ionization, or dissociation, constants defined in Equations 40a, b. (The symbol s is used for 11 sol vent".) (40a) (40b) If Equation 38 is written for the indicator, and if log (H-1) is added to each side of the equation, we may combine the resulting equation with Equation 40b to get Equation 41. PK aI = PK SaI •• Ho .. log (H.) ( 41) Since pKar is constant and pK: 1 is also constant, subject, perhaps, to small concentration effects, Ho ♦ log (H~) will be a constant. If (ff') is calculated from the total perchloric . or BH. formed acid present, corrections being made for any IH but no correction taken for molecular HCl04, then if Ho-flog (Ir) remains constant over a range of concentrations, the inference may be drawn that the perchloric acid is fully ionized. Table 25 shows for four determinations the constancy found for Ho• log (H4 ) calculated in this way. These runs were deliberately chosen to represent extreme conditions, for -3.77 and -3.95 are the smallest and largest values determined for Ho~ log (H*) and -3.30 and -1.03 are the lowest and highest values of log (H*) found in any runs during this study in which no base other than the indicator was present. In the determinations with urea very low concentrations of acid were employed. The lowest value of log (H~) was -4.12 and in this solution Ho• log (H 4 ) was -3.87. Similarly, in some runs with acetonitrile some very high acid concentrations were employed. The highest value of log (H4 ) was -0.85 and the corresponding value of Ho • log (H+) was -3. 82. Thus the sum Ho• log (H.;) was found to be constant, with a maximum and random variation of 6%, over a four thousand-fold range of concentration of perchloric acid. The calculations were made on the assumption th8.t there is no interaction between the solvated proton and perchlorate ion, and this assumption was thus found to be justified. Even the slight variation found in the value of the sum (Equation 42) may be readily explained and corrected. -so{42) It w~s shown in Table 25 that the ionization constants of the indicators, pK!I, show some variation. For any one base, of course, pK!r should be constant; and the explanation offered for the variation is that other bases are present. In the case of the 4-chloro-2-nitroaniline, the other base is probably an oxidation product, since the amine used in 541a had been exposed to air for some time and was noticeably darker than the product used for 561a. In the case of the 4,6-dichloro- 2-nitroaniline the other base was water. The more concentrated perchloric acid used in 561b had been made up with undistilled acetic anhydride and the test solution had an unknown but rather large amount_ of water present. If the ratios of the bases present are not varied too greatly, the systems are still acceptable for acidimetric determinations, since each base responds separately and in a quantitative manner to the acid present. Because the contaminants do not have the same basic strength as the indicators, however, the apparent basicity of the indicatcr solution, measured as pK: , will be diff erent 1 from the value found in an uncontaminated solution. ing to the change in pK 8 aI Correspond- a change should be made in the value of PK 8 r, or else Equation 41 and, therefore, Equation 38 will no longer be valid. Although no such correction could be made in the original calculations, it was noted that Ho• log {Ir') w2s reasonably constant as was seen in Table 25. -0.91 -0.91 -3.67 -3.67 Chloronitroaniline DichloronitroQniline Dichloronitroaniline 568a 561b 561d pKaI Chloronitroaniline Indicator 541a Run .; 0.28 0.15 3.00 2.88 'Ks pa.I -2.92 -2.80 -0.61 -0.48 Ho -1.03 -1.03 -3.30 -3.29 log (Ir) The Constancy of Ho• log (H) in Indicator Measurements. Tsble 25. -3.95 -3.83 -3.91 -3. 77 Ho • log (H') I I I-' CJ) -82- A correction can be applied to the value of pKaI for the different indicators, however, by combining Equation 42 and 41 to give Equation 43. (43) New values of pKaI for each indicator s olution can now be calculated from the observed values of pK!r as will be done in Table 26. Rather than determining the values of pKaI from Eque.tion 43 and substituting -the result in Equation 39, these equations may be combined to give Equation 44. It is from Equation 44 that the strengths of weak bases in acetic acid may most readily be calculated. s (44) : pKaB ... 3.88 2. Detailed method of calculation. of the method. The limitations Stock solutions of the indicators were prepared in acetic acid in concentrations of ab out 0.0005 M. (The acetic acid used here and in all subsequent work was purified by distillation and had a melting point of 16.6°). The absorption curve of one solution of each indicator was taken in order to find an absorption peak for the compound in this solvent. Readin gs of the test solutions were usually made at the appropriate peak in order to permit using minimal amounts of the indicator. -83- For each batch of indicator solution the optical density was specifically determined by reading known dilutions of the solution at the absorbtion peak (or, occassionally, at some other specified wave length). Multiplication by the appropriate dilution factor gave the calculated optical density of the stock solution. Dilutions were so chosen both for these neutral solutions and for test solutions containing perchloric acid so that the observed optical densities were usually between 0.200 and 0.400. Different dilutions of the same stock indicator solution gave the same calculated optical density. (When the indicator was a strong enough base to react appreciably with the solvent, a dependence of the calculated optical density on the concentration was observed. For these indicators, the optical densities of the stock solutions were calculated from dilutions containing aniline or potassium acetate. In such solutions, different pilutions of the same stock indicator solution gave the same calculated optical density.) In this way Beer's law was found to apply in the 0.200 to 0.400 range in which observations were made. s The indicator constant, KaI' was determined by mixing known amounts of indicator solution, perchloric acid solution, and (usually) extra acetic acid. The optical density of this test solution was then determined and multiplied by the dilution factor of the indicator solution in this particular test solution. That is, if 5 ml. of indicator solution were made up to a total -84- of 28 ml. of acid-indicator mixture, the optical density . observed would be multiplied by 28/5. The resultin g value could then be comparedwiththe calculated optical density of the stock indicator solution. The differen ce between the figures would be ascribed to the production of the colorless . IH+ species, · • The optical d e n sity of one stock solution was 2.742 and that of a test solution containing perchloric acid was 2.344. The ratio (I!f)/(I) was, therefore, equal to (2.742- 2.344)/(2.344). The concentration of IIr' in the solution was calculated from the k nown concentration of the s t o ck indicator solution, the d il uti on factor, and the ratio of the loss of optical density to the optical density of the indicator solution. In t h e above case, the concentration of the stock indicator solution was 0.000493 M. solution was, accordingly, The concentration of IH~ in the test (0.000493)(5/28)(2.742-2.344)/(2.742). The concentration of free acid in the so lution was determined I by calculating the stoichiometric quantity of perchloric acid present and subtractin g fr om t h is fi g ure the concentration of IW. The concentration of free acid was then divided into the ratio (IW)/(I) to give t h e value of the indicator constant, s Kar• No corrections were made for the effect of the ionic strengths of t h e solutions on the observed values of the ionization constants of either the indicators or the weak, colorless bases. Some justification is found for this simplification in the fact tha t the basic constant f ound for acetonitrile seemed -85- to be independent of the concentration of perchloric acid over a 200-fold range of concentration of the acid (Table 29). s The basic constants of colorless bases, -KB' were determined . a from data on the optical densitiesof soluti ons containing known amounts of indicator, perchloric acid, and colorless base. In such a s oluti on the valuesof (IH*) and of the ratio (IH .. )/(I) were determined as in the solutions without colorless base. The value of the indicator constant now being known, the concentration of free acid could be calculated by multiplying the ratio (I!i4)/(I) by K!r• In these solutions the material balance for the perchloric acid is shown in Equation 45. (45) The value of (Ht), .the c oncentration of the total perchloric acid added, was known,. and (IH") and as shown above.· (BH4 ). Or•) could be calculated Equation 45 could therefore be solved for The material balance for the colorless ,base in this solution is shown in Equation 46. The value of (Bt), the total concentration of the weak base in the solution was known, . and (BH) was calculated from Equation 45. The value of (B) could accordingly be derived from Equation 46. • s and ( Hy) all known, KaB was readily calculated. could then be used to find pKaB· With (B), (BH~), Equation 44 -86- The successive subtractions needed for these calculations limit the accuracy of the method. If (H~) and (Ht) are too nearly equal, the subtraction to g ive the value of (BH 4 ) will not give accurate results. Of course, more colorless base may be added to make (H~) smaller; but if (B) is made too large, the nature of the solution will change, and it is not certain that the indicator will continue to react as in more dilute solutions. If the maximum value of (B) is taken as 0.5 !, and if it is decided that at least 10% of the acid (ignoring that present as IH•) react to form BH♦, then the stren g th of the weakest base which can be measured is shown in Equation 47. The corresponding value of pKa is -4. 5. If the value of (Blr) = ( 0. 5 )( 0. 9~) = 0.22 ( 47) is too close to that of (Bt), the subtraction to give (B) will not be accurate. Of course, the amount of perchloric acid may be diminished to make (BH•) lower, but if the value of (H*) becomes too low the ionization of the solvent wi 11 no longer be suppressed and interaction of the bases with the solvent will become important. If the lowest value of (H*) considered acceptable is 0.0001 Mand if it is decided that at least 10% of the total base should remain unreacted, the strength of the strongest base which may be measured is shown in Equation 48. The corres ponding value of pK 8 is 1. 1. -87- = 3. (0.9Bt) : 90,000 (48) (Ol.Bt){0.0001) The Preparation of indicators and the properties of their solutions in acetic acid. Eastman 4-chloro-2-nitroaniline was recrystallized from boiling water and stored in a desiccator o ver sodium hydroxide. 0 It melted sharply at 116.0. were taken in an oil bath. (All melting points The te mperatures are uncorrected.) Indicator solution IA (see Table 26) was made up two months after the recrystallization and IB about five weeks later. A new recrystallization was made before preparing solution ICa. Solution ICa was found to be unstable, and new data, found after three weeks, are shown in Table 26 for solution ICb. A final recrystallization was made before preparing solution ID. In this case special precautions were taken to avoid expos- 1ing the crystals to air and this material was probably the purest obtained. It is evident that high purity is associated with a high value of pKa (enhanced basicity). To test the reversibility of the reaction cf 4-chloro-2nitroaniline with perchloric acid, two solutions of the indicator and acid were prepared. To one solution 5 ml. of 0.46 M aniline was added (the solution contained only 5 ml. of 0.005 ~ perchloric acid) and to the other 5 ml. of 0.04 M potassium acetate (this solution had 15 ml. of 0.005 M perchloric acid). The optical densities of these solutions were then determined. It was found that not c nly was the reaction reversible but the optical density of each test solution was 1% greater (after -88- adjustment with the dilution factors) than the optical density of the stock solution. Evidently, at the concentrations employed, 1% of the indicator reacted with the solvent in the readings made to determine the optical density of the stock indicator solution. This effect was suppressed by the stronger bases in the two runs just discussed. The effect would also be suppressed in all the runs made with perchloric acid. There- fore, to make comparable the readings made in tests with perchloric acid and readings of neutral solutions, all readings made in neutral solutions were increased 1%. With indicators which were weaker ·bases than 4-chloro-2-nitroaniline, this correction was not considered necessary. With indicators which were stronger bases, all neutral readings were made in solutions containing aniline or potassium acetate. To provide a base which could be used as an indicator in solutions of high acidity, the dichlorination of o-nitroaniline ' 34 was carried out in concentrated aqueous hydrochloric acid. Chlorine from a cylinder was passed through a one-liter solution containing 17 gm. of nitroaniline for about two hours~ yellow product was obtained by adding water. The The 4,6-dichloro- 2-nitroaniline was recrystallized from hot ligroin (b.p. 90-127°) to give a product with m.p. 100.7-101.0° (literature value: 101-102°). Solution IIA was used with two different batches of 0.467 M perchloric acid. The first (the data for which are shown as IIA-1) probably had a rather large amount of water in it since it was prepared from undistilled acetic anhydride. -89- The second batch of perchloric acid was made with distilled Qcetic anhydride, and the proportions of aqueous perchloric acid and acetic anhydride were calculated to give an anhydrous solution. The new data are shown for indicator IIA-2. After four and a half months, the a mine was recrystallized from ligroin and solution IIB prep a red. Although this solution was used with the anhydrous perc~loric acid, the data resemble those for IIA-1. A new, anhydrous perchloric acid solution was prepared to check these results on indicator IIB, but the indicator d a ta remained unchan g ed. The appropriate indicator constant was always used with the different solutions. The methylation of 4-chloro-2-nitroaniline was accomplished with methyl sulfate. from hot water. The chloronitroaniline was recrystallized The dimethyl sulfate was freed from sulfuric acid by shak ing with water in a small separatory funnel. Four and one half grams cf the amine (0.026 moles) and 2.5 ml. of •dimethyl sulfate (0.026 moles) were heated together in a test tube over a small flame until a sing le liquid phase was obtained. The test tube was then placed in a beaker of b oiling water for forty-five minutes. An a d ditional 0.4 ml. dimethyl sulfate was added and the heatin g continued for 15 minutes. Six ml. of 5 N sodium hydroxide were added and the product dissolved in hot ligroin {b. p. 90-107°). The material obtained on cooling the ligroin was recrystallized from hot water to give crystals which softened at 90° and melted at 106°. A further recrystalw lization from ligroin g ave 4-chloro-2-nitro-N-methylaniline with VIA VIA-2 VI B VIC vc VB VA IIIB IIIC IIID IV IIIA lA IB ICa ICb ID IIA-1 IIA- 2 IIB Indi- . cator No. ti " " - 5.81 5.53 5.52 5.52 5.75 5.40 Cone. M x 104 " 6.10 4-chloro-2-nitro-N-methylan i line 4.93 II It 5.00 II 4.27 " 11 4.27 " 2-c h loro-4-nitro-N, N4. 8 8 di me t h ylaniline 2,6-dichloro-4-nitro- N, N5.91 dimethylanili n e II II 6.78 If 5.51 " 3.18 2,4-dinitro-N, N-diethy lani li n e II It 1. 27 II 2.96 " 2.78 " II ti It 4,6-dichloro-2-nitroaniline It ti II II " It 4-chloro-2-nitroaniline Indicator Name 375 370 437 408 410 15,700 15,700 16,700 15,400 4640 4460 5560 5620 5830 5780 4470 4700 4840 4620 4930 4540 very s:nall " II -3.67 " " "tt -0.91 Absorp. Molar Extinct. pK I Coefficient at a Max. Lit. m Absorp. Max. Properties of Solutions of Indicators in Acetic Acid. Table 26. pKaI -0.97 -1.00 -0.91 -0.91 -0.87 -3.73 -3.61 -3.74 -1.37 -1.48 -1.46 -1.47 0.193 -0.17 0.172 -0.11 0,145 -0.0 4 0.0655 0.30 0.0667 0.30 0.0544 0.38 0.0725 0.26 1. 22 1. 33 1.08 1.08 0.97 701 537 725 3.08 3.96 3.78 3.85 x 10 3 Eq. 43 Ks aI I '°I 0 m.p. 106.0~106.5°. A melting point of 108° has been reportea. Analysis showed, calc'd. for c7 H7 o2 Ncl: C, 45~05%; H, 3.78%; found: C, 45.04%; H, 4.08%. Several attempts to make the di- 35 methylaniline by subsequent methylation of the monomethylaniline were unsuccessful. The first indicator solution to be prepared from the chloronitromethylaniline was indicator IIIA. A redetermination of s its optical density and Kar after three weeks showed that no change had occurred in solution. Solution IIIB was prepared six weeks later and showed tha t the crystals had undergone a considerable loss of basic strength. Successive recrystal- lizations fro~ ligroin and ethyl alcohol (solutions IIIC and IIID) gave solutions similar to solution IIIB rather than IIIA. The crystals from ethyl alcohol melted at 106.0-106.5°. Because none of the indicators studied by Hammett or Hall were monomethy lanilines, a test was made to show that the I aniliniurn ion of this indicator wu s colorless and that its formation was reversible. Three solutions were prepared as shown in Table 27, and their optical densitie~ read at 437 mr. In addition the optical densities of two of the solutions were read at several wave lengths as shown in Table 28. Table 27. The Indicator Properties of 4-Chloro-2-nitro-Nmethylaniline. Volumes are in ml. Run III D 0.000427 M HOAc No. 616a 616b 616c 3 3 3 20 15 15 HCl04 0.005 M HC104 Optical Density 0.525 M 437 mt- 5 5 0.322,3 0.253,4,2 0.013 -92- Table 28. The Optical Densities of .S olutions of Indicator IIID at Several Wa ve Lengths. mf Run. No. 616a Run Wo. 616b Ratio 437 420 400 380 360 340 0.322 0.289 0.188 0.093 0.036 0.014 0.253 0.225 0.148 0.070 0.029 0.013 1.27 1.28 1.27 i.3 1.2 1.1 s From Runs 616a and 616b, the value of K81 was found to equal 0.00390. (This result was one of two avera ged to give the 0.00385 in Table 26). s Using the value Kar: 0.00390, the optical density for Run 616c was calculated to be 0.011, a good check with the observed 0.013. A larger concentration of anilinium ion is present in 616c than in any normal observation with this indicator. If the color calculated for the free base corresponds this well to the color observed, the interference ,due to any color from the conjugate acid, IH", must be very small. Confirmatory evidence is offered by the data shown in Table 28. The ratio of optical densities of Runs 616a. and 616b at different wave lengths should be a constant in the absence of interference from the conjugate acid. The ratio is indeed found to be constant down to 400 mr- and perhaps past 380 m fl'• One pellet (0.14 g.) of sodium hydroxide was added to the soluti on of Run 616c. The next day the optical density of the solution was found to be 0.326 at 437 mt--, demonstr~ting perfect reversibility of the acid-indicator equilibrium. In fact, as with chloronitroaniline, the color returned to an optical density 1% greater than anticipated. (It was expected to equal the density found for Run 616a). In the present case, it was decided that no correction need be applied to the readings of optical density of chloronitromethylaniline in pure acetic acid. The neutralized solution had st ood overnight before the reading was made. Only one solution was available to supply data for such a correction, and a spread of 1% between duplicate solutions was comm.on. Finally the chloronitromethylaniline was a weaker base than the primary amine and would not be expected to require such a correction. The chloronitromethylaniline was used as an indicator without trouble in solutions containing aldehydes and ketones. {Chloronitroaniline was found unsuitable for use with carbonyl compounds.) Even the secondary amine, however, gave strange results in solutions containing hexyne and mercuric acetate, the optical densities dropping below values which could be explained by any known reaction. The synthesis of 4-chloro- 2-nitro-N,N-dimethylaniline was, therefore, undertaken. The successful nitration of dimethylaniline was accomplished by 36 . 0 Tingle and Blank , who secured crystals with m.p. 162-165 by the action of nitric acid on the tertiary amine. Acetic acid was used as the solvent and oxalic acid was also present during the synthesis. The authors, however, claim no special advantage nor do they give any reason for the presence of oxalic acid. They make reference to an older paper in which -94- the nitration was accomplished in acetic acid with no oxalic t 37 aci. d presen. In the attempted syntheses without oxalic acid, four different, colored products seemed to be formed - a yellow one (desired), a green one (nitroso?), a red one, and, sometimes, a brown one. Dimethylaniline was dissolved in · acetic acid at various concentrati ons. and with swirling. Nitric acid was added slowly No advantage was found in using fuming nitric acid, which often led to decomposition. At least the stoichiometric quantity of nitric acid had to be used or no product was obtained. After standing overnight, water was added and the crystals obtained by filtration. Even in the most successful runs (15 ml. of aniline and 8 ml. of concentrated nitric acid in 400 ml. of acetic acid, for example) the product was quite green. The material was recrystallized from strong hydrochloric acid by dilution (always followed I by washing with aqueous ammonia) and then from water or ethyl alcohol. The green color persisted through all recrystalliz- ations, though a melting point as high as 163-165° was obtained for one product. It was found that if the crystals were taken up in warm pyridine, the addition of 30% hydrogen peroxide discharged much of the green color. Attempted syntheses in the presence of sulfamic acid gave no product even after standin a week. Tingle and Blanck's method was found to be much more successful. To 100 ml. of acetic acid were added 36 gm. -95- anhydrous oxalic acid (which did not all dissolve), 13 ml. dimethylani line, and 7 ml. of nitric acid. Frothing and red color were nobed . in the solution when only half the nitric acid had been added, but best results were obtained if the rest was added anyway. The best run was made in the presence of 1/2 gm. of urea and 5 ml. of acetic anhydride. The reaction was quenched by adding water after about one minute Qnd the product worked up in the usual way. The purest material was obtained from an unsuccessful attempt at chlorination in a solution of acetic acid and concentrated hydrochloric acid. Dropping concentrated potassium permaganate on concentrated hydrochloric acid did not generate enough chlorine to effect a chlorination, but beautiful yellow needles of p-nitrodimethylaniline were recovered, m.p. 163-165°. The chlorination of p-ni trodimethylaniline 38 was ac,c wmplished by passing chlorine from a cylinder through a solution of 4 g. of the amine in 8 ml. of acetic acid and 16 ml. of concentrated hydrochloric acid. After 2 gm. (85%) of chlorine had dissolved, the chlorine w2s turned off and stirring continued for 15 minutes. The product was obtained by adding ice, sodium bi- carbonate, and sodium hydroxide. Successive recrystallizations from ethanol and ligroin gave yellow plates of 2-chloro-4nitro-N,N-dimethylaniline, m.p. 73-75°. (Drake reported 76-77°.) Analytical results were: calc'd. for c8 H90 2N2 Cl; Cl, 17.67%; found: Cl, 17.85%. -96- Indicator IV was prepar ed from this material. The maximum optical density was at about 370 mt--; but a test, analagous to that demonstrated in Table 28, indicated that interference from the color of the Irr' ion was noticeable at wave lengths shorter than 390 mf-. Readings were therefore made at 410 mf'-. It was found that in a solution only 0.000536 Min perchloric acid, 98% of the color of this indicator was discharged. Since the use of perchloric acid in concentrations much lower than five thousandths molar were not contemplated, no further work was done with this indicator. The preparation of dichloroni trodimethylanilirre next undertaken. was The p-nitrodimethyl aniline was dissolved in acetic acid and hydrochloric acid as before, the intention being to pass in a considerable excess of chlorine. However, dark colored material appeared soon in the solution so the reaction was stopped and the monochlor product isolated and recrystallized I from alcohol, m.p. 65-72°. These crystals.were then chlorinated further, the solution remaining a pale yellow. It was noted that the solution became warm during the chlorination, and so the flow of chlorine was continued until the heat effect was no longer noticeable. The crystals of 2,6-dichloro-4-nitro-N, N-dimethylaniline were obtained by pouring the solution onto ice and were recrystallized from et·hyl alcohol. 105-107° (literature value; 104-105° 38 .) They melted From this material indicator solutions VA,B, and C were prepared. The absorption maximum was found to lie at about -97- 380 mf, but by the techniques already described, it was found that the conjugate acid was not colorless in this region. Read- in g s of test solutions were, therefore, made at 420 m~. The solutions were unstable. 1.41 to 1.56 in nine days. The optical density of VC changed from Moreover, these solutions, too, gave strange results in solutions containing hexyne and mercuric ~cetate. These new data made it seem possible that the chlorine atom or atoms, rather than the amine hydrogen atom (of which there are none in indicator V) wer~ responsible for the anomalous results obtained with indicators III and Vin solutions containing hexyne and mercuric acetate. The synthesis of 2,4-dinitro-N,N-diethylaniline was, therefore, undertaken. Some of the indicators investigated by Hammett were polynitro bases, and they were reported as suitable for this work. Ten grams of Eastmann 2,4-dinitrochlorobenzene (0.05 moles) were dissolved in 20 ml. of butyl alcohol. (0.10 moles) of Eastman diethyl amine were added. slowly became warm and then quite hot. with a reflux condenser. Ten ml. The so.lution The flask was fitted After the spontaneous heat of reaction had subsided, in about five minutes, the solution was refluxed for 15 minutes. Water was then added and the crystals recrystal- lized twice from ethyl alcohol. 78.8~79.2°. They softened at 70° and melted Some of the material was then recrystallized from strong hydrochloric acid by the addition of sodium bicarbonate. Recrystallization from ethyl alcohol gave crystals of 2,4-dinitroN,N-diethylaniline, m.p. 78.5-79.0°. value was so 0 39 • The previously reported -98- Solution VIA was prepared from this material after ten days and solution VIA-a was made by making 20 ml. of VIA up to 50 ml. with glacial acetic acid. Four days later solution VIB was prepared, and showed a marked increase in basic strength. For solution \~Can entirely new preparation of the amine was used 'and dissolved in acetic acid only three days after being prepared. Unlike the other indicators this material seems to gain in basic strength on standing in air. 4. The purification of the colorless bases and the determination of their basic strengths. Purified dioxane was generously furnished by M. Pierre Leroux. It had been refluxed for five hours over sodill..rrl hydroxide and for five hours ever sodium before distillation. 5 Its refractive index, n~ , was 1.4195. value is 1.4220 at 20° 40 .) (Literature Two solutions in acetic acid were prepared; one, 0.926 Min dioxane, was used for the first determination (Table 29); and the other, 1.40 M, was used for the other three determinations. A definite concentration effect is to be noted in the calculated basic constants for dioxQne. The best value for pKa is about -4. Eastman acetonitrile was dried with phosphorus pentoxide and distilled. The observed index of refraction, ~ 5 , was 1.3409. The literature 41 values are 1.34604 et 15°; 1.33934 at 30°. Linear interpolation gives 1.3413 at 25°. The first four reaaings in the table were made with a 1. 90 M solution -99- prepared from this material. The next five readings were made from a similar solution prepared from newly purified acetonitrile. The best value for pKa is -4.2. Hammett and Deyrup found -4.19 for propionitrile. 33 b, 42 Dr. Judd Nevenzel was kind enough to furnish some .E,-toluenesulfonamide which had been recrystallized from ethyl alcohol and found to melt from 137-138°. in acetic acid. The material was not very soluble An attempt to dissolve 1.50 gm. in 25 ml. of acetic acid failed, but 1.026 gm. did dissolve to give the 0.248 M solution used in the three determinations shown in Table 29~ The best value for pKa is -3.2. Laboratory distilled water was employed for determining the basicity of water. Since only 1% of the water reacted in the first determination and 3% in the third, the satisfactory agreement between these runs serves to demonstrate that no strong base (such as ammonia) played an appreciable part in 'these determinations. The best value for the pK about -2.3 5 • Ham~ett and Deyrup found -3.72. . k nown. 33b,42 f or th e di screpancy is of water was a No explanation Recrystallized acetanilide was generously furnished by Dr. Fern Mitchell. It was dried in a vacuum over phosphorus pentoxide at the temperature of boiling ethyl alcohol for two days and was then found to melt 113.0-113.3°. One solution of the amide was used for the first and third determinations; another, for the second; and still another, 4 months later, for the last three. At the time the l a st three runs were made, 1/5 1/5 5/23 1/6 1/7 1/4 5/28 2.55 2.452 2.836 2.60 2.742 ICb IIA-1 ID IA IIIA p-Tolueneso lfonamide Wa ter 1/4 1/5 5/28 1/8 5/28 1/4 1/5 1/6 1/5 2.60 IB Acetonitrile 1/5 1/5 1/5 5/23 2.836 ID Dioxane B&se 0.200 0.263 0.202 0.0319 0.0328 0.0300 0.02'66 0.362 0.309 0.320 0. 326 0.250 0.369 0.465 149 44.2 28. 8 100 85. 7 50.0 119 1.67 2.14 2.50 0.893 -2.38 -2.40 -2.36 .;,2 . 31 -3.18 -3.30 -3.19 ..,4. 21 -4.35 -4.30 -4.28 -4.26 - 4 .21 -4.22 -4.05 -4.11 2.15 2.96 2.66 2.50 2.42 2.14 2.17 1.47 1.69 0.400 0.318 0.355 0.235 0.3325 0.361 0.4085 0.379 0.442 951 391 , 680 476 680 951 761 633 381 1.250 1.000 0.536 0.625 0.536 1.250 186.8 78.2 93.4 1.00 0 1.000 1.087 -4.32 -4.31 -4.24 -4.12 2.74 2.66 2.30 1.73 0.313 0.325 0.3035 0.315 pKaB s KaB 556 840 280 183 - Basic Strength 1.000 1.000 1.00 0 1.087 - Indic2tor Test Solution Base Solution Optical Indicator HCl04 (Table 26) dilution total cone.total cone.Density M x 103 M x 103 Observed Opt.Dens. factor No. The Determination of the Strength of Weak Bases. Table 29. I 0 0 I ..... 2.60 2.742 2.67 2.67 IIIA ICa ICa Acetamide Urea Acetanilide IB Base 1/9 1/9 1/9 5/28 5/33 1/8 1/6 5/23 1/8 1/6 5/23 1/5 1/4 1/5 1/8 1/6 1/6 5/28 0.556 o.556 0.556 0.536 0.455 0.48 0.44 0.51 0.62 0.57 0.279 0.2775 0.4445 0.3665 0.2725 0.791 0.791 0.648 0.635 0.443 0.04478 0.04436 0.04408 0.04316 0.04355 0.000166 -0.10 0.000146 -0.04 0.000125 t0.06 0.000169 -0.10 0.000174 -0.11 0.000113 0.07 0.3125 0.384 0.435 0.307 0.363 0.4135 1.802 1.202 0.941 1.470 0.982 0.769 0.625 0.833 1.087 0.625 0.833 1.087 pKaB 0.00121 -0.96 0.09103 -0.89 0.00153 -1.06 0.00123 --0.97 0.000885-0.83 o. 00126 -0. 98 0.00111 -0.93 KaB s Basic Strength 0.3865 0.400 0.3315 0.2375 0.395 0.377 0.394 1.86 0.820 0.620 0.410 0.940 0.470 0.302 1.00 1.25 1.00 0.625 0.833 0.833 0.893 Indicator Test Solution Solution Indicator HClOi Base Optical (Table 26) dilution tota conc.tot~l cone.Density No. Opt.Dens. factor M x 103 M x 103 Observed Table 29 (Continued) I I-' 0 I-' I -102- the acetanilide was found to melt from 113.0-114.5°. The test solution for the fourth determination was prepared by diluting an aliquot from the second test solution 1:1 with acetic acid. The satisfactory agreement between these runs serves to demonstrate the reversibility of the reaction between acetanilide and the acid as well as confirming the reversibility of the indicatoracid equilibrium. The values of pKa average about -0.9. · Hammett and Deyrup reported values of -1.89 and -1.74. 33b 42 ' Acetamide (Merck, reagent) was recrystallized from freshly distilled isopropyl ether and dried in a vacuum desiccator over phosphorus pentoxide. The needle-like crystals, very dry and crisp, melted 80.0-80.8°. Two acetic acid solutions were pre- pared, one used for the first three determinations and the other for the last three. Except for the fifth determination it will be noted that with higher ratios of perchloric acid to amide the apparent basicity of the amide increases. This observation may be associated with the ability of two moles of the amide 43 to react with one mole of strong acid. The best value for pK with the acid in excess is probably about 0.1. 44 results in the literature average about -0.5. 8 Previous Urea (Eastman) was recrystallized from absolute ethyl alcohol and dried in a vacuum desiccator, m.p. 132.2-133.5°. Three different solutions were prepared. One was used for the first detennination; one, for the second and fourth; and one, for the third and fifth. The ratio between acid and base was -103- not varied enough to make any trend recognizable. The values of pKa averaged sbout 0.5. Previous values in the literature 44 average about 0.1 , The greater basicity observed for area and acetamide in these measurements as compared with older constants measured in water may be due in part to the high concentrations which had to be used in aqueous solutions to measure the basic strength. In addition to these colorless bases whose strengths were determined, a number of colorless compounds were investigated for which values of pK 8 could not be found because they lay outside the range from -4.5 to 1.1. As a test of the method heptane (a highly purified sample generously furnished by the Philips Petroleum Co.) was dissolved in acetic acid and used as a base in a several determinations with indicators IB and IIIA. s In concentrations ranging up to 0.9 M, Kb was found to equal 0.00 well within the experimental error. Benzophenone, ethylene diacetate, and cyclohexene all failed to show measurable basic strength. Dibutyl ether, acetone, and acetophenone were on the borderline. The readings made on solutions of these materials indicated a small degree of basic strength, but the differences from blank readings were too small to be safely distinguished from .normal dispersion. Diphenylamine and o-chlordaniline were found to be too strongly basic for their basic strengths to be measured by this method. -104- C. The Formation of Complexes with Mercuric Acetate by Hexyne and Hexeny1 Acetate. The experience gained from the indicator studies on colorless bases was now applied to the study of the mercury complexes. That Equations 26 and 27 were written without acid was simply due to lack of knowledge. If acid did form part of the complexes, the measurement of the acidity of the medium might permit calculation of the equilibrium constants of the complexes. First, the effect of hexyne and mercuric acetate separately on the acidity of perchloric acid solutions was studied. Using indicator IIID, it was found that 0.30 M hexyne had no effect on the acidity of a 0~0017 M solution of perchloric acid in acetic acid. Two later tests with indicator VC con- firmed the knowledge that hexyne, by itself, Would not effect the acidity of acetic and solutions of perchloric acid. Mercuric acetate, however, was found to be somewhat basic. The probable reaction of mercuric acetate with the solvated proton in acetic acid is shown in Equation 47 and the equilibrium constant, Kb, is defined in Equation 48. (47) s ¾• (HgOAc•) (Hg(OAc)2)(H 4 ) (48) Two similar tests of the value of K~ were made with indicator IIIB and perfect agreement was obtained as shown in Table 30. -105- A third test with indicator IIID gave good agreement, and the value of K~ is probably quite close to 70. (In this case K~ instead of Ks is listed, since it seems far-fetched to c~ll a mercuric monoacetate ion a conjugate acid. However, despite the slight difference between Equations 44 and 47, the calculations for Table 30 are virtually identical with those for Table 29~) Table 30. The Basic Strength of Mercuric Acetate in Acetic Acid Test Solution Indicator Indicator HCl04 HgOAc2 Optical Base Solution Dilution total cone.total cone.Density Strength M x 10 3 M x 103 Observed Ks No. Opt.Dens.Factor b IIIB 2.804 5/28 0.535 14.3 0.480 74,74 IIID 2.455 5/28 1.783 2.14 0.3135 66 The acidity of solutions containing both hexyne and mercuric acetate in addition to perchloric acid were then investigated. Because of the yellow by-product formed in such solutions, blank runs, without indicator, had to be run to correspond to each test solution containing indicator. The optical densities of the blanks - taken at the same wave length and after the same time interval as for the test solutions were subtracted from the optical densities of the test solutions. The difference, called the "indicator optical density", was then inserted in the usual calculations. Indicator IIrD was the first indicator to be tried in solutions containing mercuric acetate and perchloric acid. -106- The indicator optical .density was found to drop precipitously to a value so low th at it could not be explained even on the assumption th at no perchloric acid was being bound by the hexyne or mercuric acetate or any complex of the two. This observation would be expected if the basic nature of the indicator were being diminished without any corresponding diminution of the light-absorbing quality of the nitro groups. The same type of results were obtained with indicator VC. It was with indicators VI A and VI B (2,4-dinitro-N,N• diethylaniline) that the nature of the hexyne-mercuric acetate complex was finally elucid~ted. ments are shown in Table 31. The conditions of the experi- (Some data on the equivalent blank determinations were given in Table 15, runs Cl7-C21.) An attempt to extrapolate the indicator optical densities back to zero time gave 0.366, 0.360, 0.227, 0.400, and 0.165 as the best values for the five runs. From these figures the con- centrations of Irr', of free ·If, and of combined acid (exclusive . of IH) were calculated in the usual manner. (See Table 32.) The hypoth~sis was then made that the mercuric acetate and the perchloric acid combined in a 1 to 1 ratio. On this basis the amount of free mercuric acet ate could be determined. It was then possible to calculate the concentration of mercuric mono~ acetate from the data of Table 30 and to allocate the combined acid - part to the mercuric monoacetate and part to the hexyne- mercuric acetate complex. The additional hypothesis was then made that there was one molecule of hexyne in each molecule of complex. The amount of free hexyne could then be calculated. 694a 694b 696a 699c 704a No. Run 694a 694b 696a 699c 704a No. Run 4.97 4.97 4.97 4.97 4.98 0.400 0.600 0.606 0.467 0.750 0.366 0.451 0.533 0.387 0.667 0.800 0,500 0.182 0.667 0.312 Table 32. 0.424 0.530 0,643 0.212 0.331 (0.370) 0.373 0.227 0.398 0.165 0.376 0.385 0,227 0.401 0.168 0.394 0.225 0.388 0.166 0.395 0.228 0.387 0.164 o. 386 0.115 0.163 o.375 0.209 0.456 0.247 0.286 0.157 0.171 0.208 0.004 0.002 0.001 o. 007. 0.003 0.424 0.530 0.643 0.212 0,331 0.177 0.244 0.486 0.041 0.125 Hexyne Free Total 0.800 0.500 0.182 0.667 0.312 0.549 0.212 0.024 0.489 0.101 Hg(OAc)2 Free Total 2.2 X 107 3.4 X 10 7 3.6 X 107 4.1 X 107 3.6 X 107 Equation 49 K Computation of the Equilibrium Constant of the Hexyne-Mercuric Acetate-Perchloric Acid Complex. All concentrations are M x 10 3 1/5 1/4 10/33 1/3 10/32 Perchloric Acid Combined Free Total-IH" Complex HgOAc" VIB VIA VIA VIA VIA 0.383 0.400 0.231 0.390 Test Solution Indicator Hexyne Hg(OAc)2 Indicator Optical Density No~ Optical Indicator HC10 4 Density Dilution total total total _ _ Factor M x 103 M x 103 M x 103 1 min. l½ min. 2 min. 3 min. 4 min. Observations on the Hexyne-Mercuric Acetate-Perchloric Acid Complex. Table 31. I ....J 0 I r-' -108- From the amount of complex and the amounts of free acid, mercuric acetate, and hexyne, the equilibrium constant for a 1 to 1 to l complex was calculated according to Equation 49 • . Complex K: (C5H1o)(HCl04)(Hg(OAc)2) (49) The good agreement obtained in the widely diversified runs lends credence to the hypotheses used in the calculations. Of course, the data show only the ratio in which the constituents enter the complex and give no information concerning its structure. The most reasonable theory seems to be that.the role of the acid is to combine with one of the acetate radicals, leaving the complex as a hexyne-mercuric monoacetate cation. is embodied in Equation 50. This suggestion Regardless of the structure of the complex, it is concluded that a 1 to 1 to l complex is formed from hexyne, mercuric acetate, and perchloric acid. The best value for the equilibrium constant is about 3.5 x 107 moles" 2 liters 2 • One surprising feature of the data in Table 31 should be pointed out. The value of the equilibrium constant depends directly on the amount of excess hexyne in solution. The amount of this excess is believed to be rapidly changing as the hexyne reacts with the solvent to form hexenyl acetate. However, only in the first two runs does there appear en initial trend of any important size in the indicator optical density. -109- Similar determinations were then made with solutions containing hexenyl acetate, mercuric acid, and perchloric acid. The data and the results of the calculations are sho wn in Table 33. The fi gure shown for the concentration of the complex in each run is simply the concentration of the acid not otherwise identified as fre·e or present as IH.;. At the concentrations used, no correction had to be made for mercuric monoacetate. The values of K were calculated, as in Table 32, on the assumption that a 1 to , 1 to 1 complex was bein g formed. Many of the results lie between 3.2 and 6.0 x 107 , but the discrepancies seem to be significant. When the perchloric acid is in con- siderable excess over the hexenyl acetate and mercuric acetate (Runs 737b and 737c), the value of K is smaller than usual. When the hexenyl acetate and mercuric acetate are present in markedly unequal amounts (Runs 745c and 745d), the concentration of complex formed seems to nearly equal the concentration of the reagent present in smaller quantity. If the equilibrium constant were calculated in the usual way _for these tests, the values would be very large or infinite. The hexenyl acetate has three sites avail able for possible complexing with the mercury. It seems reasonable tha t the equilibrium involved should be more complicated than with mercuric acetate and hexyne. 0.312 0.214 0.116 0.100 0.136 0.156 0.139 0.139 0.132 0.132 0.104 0.104 0.109 0.119 0.119 3/16 3/14 2/13 . 2/15 2/11 3/16 1/6 1/6 3/19 3/19 1/8 1/8 3/23 1/7 1/7 737b 737c 740a 740b 740c 739a 739b 742e 741a 742b 741b 742c 742f 745c 745d 0.148 0.163 0.182 0.263 0.216 0.148 0.219 0 -. 219 0.125 0.125 0.099 0.099 0.172 0.378 0.113 - Run No. - Hexenyl In d icator HCl04 Acetate Dilution total total M x 10 3 M x 103 Factor 0.1560.179 0.192 0.167 0.136 0.156 0.139 0.139 0.079 0~079 0.063 0.063 0.109 0.071 0.238 0.304 0.4.08 0.572 0.268 0.318 0.353 0.186 0.316 0.377 0.356 0.408 0.378 0.321 0.319 0.272 0.268 0.304 0.408 0.572 0.378 0.355 0.188 0.319 0.377 0.356 0.1 90 0.322 0.378 0.357 0.409 0.379 0.356 0.353 0.320 0.320 0.274 0.271 0.305 0.410 0.572 0.192 0.324 0.379 0.358 0.409 0.379 0.357 0.355 0.320 0.320 0.275 0.272 0.307 0.412 0.572 H ( ) Indicator Optical Density g OAc 2 total Initial M x 103 Est. 1 min. 2 min. 3 min. ' Calculated Quantities 0.88 3.8 0.031 0.067 0.109 ? ? 5.0 6.0 3.9 4.9 4.6 5.2 3.2 1. 5 0.014 0.027 0.031 0.037 0.044 0.052 0.046 0.047 0.041 Complex K M X 103 X 10 7 Observations on the Hexenyl Ac~tate-Mercuric AcetatePer~hloric Acid Complex Table 33 . . I I-' f-' 0 I ... 111- D. The Acidity of Solutions of Sulfuric Acid in Acetic Acid. An important paradox must be pointed out in connection with the acidity of acetic acid solutions of sulfuric acid. Hall found that his and Hammett's data suggested that sulfuric acid was a strong acid in acetic acid solutions less than 0.2 M in sulfuric acid. A plot of Ho against log(H 2 S04) gave a straight line with a slope of -1. This observation is equivalent to Equation 51, which resembles Equation 42. It might, therefore, Hof log (H*): Q (51) be concluded (and Hall and Spengeman did conclude) that sulfuric acid is a strong monobasic acid in acetic acid. shows the calculation of Q. Table 34 The first two runs were made as part of the present research, using indicator IIID. runs were reported by Hall and Spengeman. The other 31 Table 34. The ~ cidity of Solutions of Sulfuric Acid in Acetic Acid H2 so 4 ,M Ho Ho • log (~S0 4 ) 0.00332 0.00909 0.00315 0.00590 0.00954 0.0103 0.0411 0.241 0.651 1.794 -0.15 -0.58 0.05 --0.30 -0.60 -0.49 -1.10 -1. 93 -2.65 -3.21 -2.63 -2.62 -2.45 -2.53 -2.62 -2. 48 -2.49 -?.55 -2.84 -2.96 \ -112- It will be observed the Q is, indeed, fairly constant for solutions in which the acid concentration is 0.2 Mor less. Instead of agreein~ with Equation 42, however, it turns out that Q has a value of about -2.55. Sulfuric acid solutions, then, are much less acidic than solutions of perchloric acid. 1 3 In fact, a solution of sulfuric acid must be twenty (10 • ) times as concentrated as a solution of perchloric acid to give a solution of equivalent acidity. There is no evidence that makes it necessary to revise the theory tha t perchloric acid is fully ionized. It is necessary to postul;ate, however, either that sulfuric is rather slightly ionized or that some process other than ionization into proton and bisulfate ion uses up molecular sulfuric acid. Because the surprisingly low acidity of the sulfuric acid solutions had no known explanation, it was decided to use perchloric acid in the kinetic studies which were now to be resumed. -113- VI. Reaction of 3-Hexyne and 3-Hexenyl Acetate~3 with Acetic Acid. Final Results. A. The Mechanism of the Reaction of 3-Hexyne. Runs 101 to 116. These determinations were made after the decision had been reached to use perchloric acid for the final kinetic tests but before the hexyne-mercuric acetateperchloric acid complex had been successfully measured. The hexyne which was present in these runs was pipetted from a stock solution of hexyne in acetic acid which had been prepared two months earlier. Bromate-bromide titrations of known dilutions of this solution were only about 3% hi gh instead of the usual 6% (see Table 8). It is probable that about six percent of the hexyne had reacted to form hexenyl acetate during the interval. The formation of peroxides may also have occurred. Therefore, the data for these runs must be evaluated with some caution. They are shown in Table 33. No general interpretation was found which would fit all these runs. However, certain qualitative conclusions can be drawn from the data. The tapering off phenomenon was most noticeable when the concentration of hexyne was in considerable excess over the concentration of the catalysts. When the initial concentration of the hexyne was low, a larg e fraction of the hexyne reacted very quickly as in Runs llOa and 113. However, in all the runs in which successive readin~ s could be obtained, the rate was found to taper off from a zero or first order dependence on the concentration of the hexyne. Runs 105 and 109 were duplicates and a greed quite well, but Runs 110a and 110b -114Table 33. Run No. Hg(OAcJ 2 M x 10 101 5 HCl04 M x 103 5 102 5 5 105 2 2.5 106 2 2.5 109 (105b) 2 2.5 110a 2 2.5 llOb 2 2.5 111 3 2.5 Time Min. Hexyne M X 103 0 (44.2) 4 38.3 21 41 58 0 2.5 24 0 3.75 24.5 32.4 44 0 4.25 27 49 0 2.75 25 0 3.5 25 0 3 25 0 3 112 3 2.5 113 3 2.5 114a 6.0 o. 5 / 114b 6.0 0.5 116 0.6 0.5 24 0 2.75 22 0 3 22 0 3.75 ' 24 0 3.5 26 0 2.75 22.5 28. 7 27.8 (88.5) 76.7 73.3 (13.2) 12.5 7.1 5.5 (6.62) 5.10 2.35 1.20 (13.2) 12.3 7.1 (3.98) 1.61 0 ( 3. 98) 2.41 1.29 (13.2) 12.3 6.8 (6.62 5.50 3.10 ( 3. 98) 0.62 0 (2.65) 1.47 1.47 (2.65) 1.35 1.55 (2.65) 1.98 1.62 -115- are noticeably different. Some of the runs catalyzed by 0.002 M mercuric acetate and 0.0025 M perchloric acid seem to have a faster initial rate than runs catalyzed by 0.003 M mercuric acetate and 0.0025 M perchloric acid. From these runs it was concluded that catalyst concentrations of about 0.0005 M to 0.0010 M would provide rates which it would ·be convenient to determine. Runs 117-128. Following the determination of the equili- brium constant of the hexyne-mercuric acetate-perchloric acid complex, a set of kinetic determinations was made in which the perchloric acid and mercuric acetate concentrations were the same in any one run. This single concentration was varied between 0.0003 Mand 0.0017 M to include the range suggested in the last paragraph. The data are shown in Table 34. from the table. Six runs are omitted Runs 122 and 123 were made after the stock supply of hexyne had been exposed to air and before it was redistilled. Runs 124, 125, and 126 unaccountably gave precipitates. Run 124 was 0.038 Min hexyne, 0.001 Min mercuric acetate, and 0.0005 Min perchloric acid. Runs 125 and 126 had the same concentrations as Run 127. The case of Run 119 is discussed in the next paragraph. The kinetic determinations were usually made with solutions havin g a volume of 50 or 100 ml. A solution of 100 ml. 0.01 M in hexyne would have less than 0.1 ml. of hexyne, an amount too small to pipette accurately. In Runs 101 to 116 a stock solution of hexyne in acetic acid was used as an intermediate between the -116- hexyne and the kinetic solution. However~ evidence was presented that a slow reaction took place in this solution. In Run 117 and in most succeeding runs, a fresh intermediate solution was prepared for each kinetic determination. Occassionally, one such solution would be used for two runs if it was possible to start the second run within three hours after preparing the intermediate solution. The concentrations of the intermediate solutions were determined by the titration of aliquots until sufficient data were available to allow the prediction of the concentration of a certain dilution of hexyne. From the known dilution of the intermediate solution used in the kinetic solution, the initial concentration of hexyne in the kinetic soluti0n could be calculated. In Run 119 the concentration of the intermediate solution was found to change with time; and this run is, therefore, omitted from Table 34. This instability is believed to be related to oxygen dissolved in the hexyne. Mr. Dorsey has found that the rate of disappearance of unsaturation in an aqueous solution of hexyne is increased by the presence of oxygen. Run 117 w~s made with hexyne which had not been distilled for six months. in Table 34. The initial rate in this run was the largest The hexyne was then distilled f ·or use in Runs 118-121. In these determinations the rates we r e slower than in Run 117. Finally the hexyne was redistilled for Runs 127, 128, and 128s. Run 127, which had the same amount of catalyst as Runs 118, 120, and 121, was found to react much more slowly -117 ... Table 34 The Rate of Reaction of Hexyne and Acetic Acid. Catalysis by Mercuric Acetate and Perchloric Acid at 250. Run No. Hg(OAc )2 M x 10 3 HC10 4 M x 103 Time Min. Hexyne M x 103 117 0.5 0.5 118a 0.5 0.5 118b 0.5 0.5 120a 0.5 0.5 120b 0.5 0.5 121 0.5 0.5 127a 0.5 0.5 127b 0.5 0.5 128 0.33 0.33 128s 1.67 1. 67 0 3.25 30 91 125 153 0 3.5 9.5 0 3.7 17 0 3.2 14.5 0 2.5 20 0 2.75 14.5 0 2.1 5.2 0 117 0 2.1 4.2 0 2.5 5.5 ( 52} 46 33 31 32 32 (24.2) 23.8 23.3 (24.2) 23.8 23.4 (55.3) 53.7 53.5 (55.3) 54.2 52.5 (10.1} 9.7 9.1 (34.0) 34.0 34.0 (34.0) 34.0 (34.0) 34.0 34.0 (34.0) 32.2 32.2 -118- than th e e arlier thr ee runs. In fact no ch t.ne; e in unsaturation 0 seemed to t ake p lace in either Run 1 27 or 128. Jt was concluded tha t the kine tic dete~1 ina tions should be made uncier nitro g en, using hexyne whi ch h a d never been exp osed to air. It wa s furt he r d eci de d th&t if c a talyst concentr a tions of about 0.001 M we re to be emp loy e d, lower · initial concent rations o f hexyne would have to be used to prevent mas k ing a s ma ll amo unt of cha n g e with a la r-g e amount of unchan g ed materials. Runs 129 through 138. The experi ence g ained in the previous woI'i-:. wa s a pp li ed to Runs 129 through 138, from which t h e me chanism of the reaction wa s finally d educed. The runs were made un o.e r nitro g en, and.nitro g en wa s a lso used to pr otect the int erm edi a t e hexyne solution. Because of t he nwnb er of diffe rent solutions (4) wh ich had to be mixed to make the k in e tic solutions, becau se no conv en i e nt method wa s a vail a ble fo :r- t her-mostat:lng the a li quots durin g the ti me ne e ded for srur,p ling , 2 nd because previous experience had demon s tr a ted no s trone; t emperature effect on t he rate, these r v. ns were mad e c.-.t room t emperatu re. The res u lts are shown in Table 35. If just Runs 129 throubh 136 be cons i der ed fir s t (runs in which th e concen tra tion of mercuric ac e t a te equaled the concentre ti on of per ch loric c:, c id ), it wi J..l be s een t r.Ed; t he t es t s c a n Le d.ivi d ed into t h ree g ro up s on t he b asis of t h8 change in h exyn e concen t r ation in t he fir s t 2 1/2 minutes of ea ch run. ('l'he c on c enT, r e,ti on a t the e nd of 2 1/2 minutes was d et e .crn l ned -119- for each run by reading from a graph of concentration vs. time. The first aliquot was usually taken at about 2 1/2 minutes.) In Runs 133 and 136 this chan g e is about 0.000-70M; in Runs 129, 131, 132, and 135 it is about 0.00150 M; and in Run 130 it is 0.00225 M. It is immediately evident that this diversion has grouped the runs according to the amount of catalyst pr e sent and that the grouping is independent of the total concentration of hexyne. Moreover, the three concentration changes, 0.00070, 0.00150, and 0.00225 M, are approximately in the ratio l:2:3, which is exactly the ratio of the concentration of - catalysts in the respective runs. The only mechanismthat is b~lieved toaccount for these findin g s is that the reacting spe cies is a complex, formed by the hexyne with the catalysts. Only in this way can the inde- pendence of the initial rate of reaction of the hexyne from the initial concentration of the hexyne be explained. In order to determine the ratio of perchloric acid to mercuric acetate in this complex, Runs 137 and 138 were made with unequal amounts of the two catalysts. In Run 137, which was 0.0005 Min mercuric acetate and 0.0015 Min perchloric acid, the concentration of hexyne dropped 0.00105 M in the first 2 1/2 minutes. In Run 138, mich was 0.0015 Min mercuric acetate and 0.0005 Min perchloric - - acid, the concentration dropped 0.00ll0M in the first 2 1/2 minutes. These data do not fit perfectly the initial rates of Runs 129 through 136. If the complex contained one mole each of mercuric acetate and perchloric acid, the concentration of -120- Table 35. The Ra te of Reaction of 3-Hexyne with Acetic Acid, under Catalysis b y Mercuric Acetate and Hitrofeh; Perch oric Acid. Run No. Temp. 129 29 1 130 23 131 Min. Hexyne M x 103 1 0 3.5 6.5 45 (5.80) 4.02 3.92 3.48 1.5 1.5 0 3.5 31 60 1640 (5.80) 3.34 2.37 2.04 1.79 25 1 1 0 2.67 10 44 131 (5.80) 4.30 3.75 3.16 3.12 132 24 1 1 0 2.75 10 36 800 (11~60) 10.15 9.60 9.05 8.85 133a 25 0.5 0.5 0 3.1 11 46 (5.80) 5.26 4.95 4.91 (De g . ) Hg(OAc)2 M x 103 HCl04 M x 103 Time 133b 27 0.5 0.5 0 2.25 9.5 (5.80} 5.15 4.85 133c 25 0.5 0.5 0 2.25 10.2 39 (5.80) 5.10 4.91 4.78 135 25 1 1 0 2.5 9.7 45 ( 3. 48) 1.88 1.35 1.17 -121- Table 35 (continued) Run Temp. Time Min. Hexyne M x 103 0.5 0 3.5 13.7 (3.48) 2.78 2.56 1.5 0 3 11 (5.80} 4.66 4.22 3.34 1.80 No. (Deg.) Hg(OAcl2 M x 10 HClOf M x o3 136 25 0.5 137 26 0.5 39 143 138 28 1.5 0.5 0 2.25 9.5 48 (5.80) 4.75 4.32 4.36 ... 122- hexyne should have dropped 0.00070 Min the first 2 1/2 minutes in Runs 137 and 138. If the complex contained two moles of one catalyst and one mole of the other, then either Run 137 or 138 should have lost 0.00150 Min the first 2 1/2 minutes, the other losing 0.00070 M. The values actually found, 0.00105 M and 0.00110 M, suggested that the complex actually contained only one mole of each catalyst. The extra hexyne reacting might be the result of the catalytic action of excess perchloric acid (in Run 137} or of excess mercuric acetate (in Run 138) on the excess hexyne present in these runs. ("Excess" is used to mean the material not involved in the complex.) In order to determine the probable rate at which the excess hexyne would react in the presence of an excess of one catalyst, a series of determinations was made in which only one catalyst was present. The data are shown in Table 36. It is evident that the small size of the effect being investigated placed a great strain on the techniques of mixing, sampling, and titrating these solutions. The results in the three acid- catalyzed tests (Al, A2, and A3) were scattered, and only in Test A3, with a relatively high concentration of perchloric acid, was a clear-cut effect demonstrated. On the basis of the three tests, the best value for the amount of hexyne reacting in 2 1/2 minutes in the presence of 0.0010 M perchloric acid is 0.00010 M. - - If this figure is subtracted from the 0.00105 M found in Run 137, the difference, 0.00095 M, should be the initial loss of hexyne due solely to the reaction of the complex. This value is still ..;123- intermediate between 0.00070 Mand 0.00150 M, the values which - - would indicate, respectively, the inclusion of one or two moles of perchloric acid in the complex. However, anothe r effect was operatin g to make Run 137 different from Runs 129136. Run 137, with the excess perchloric acid, showed the taper- ing-off phenomenon much less than the earlier runs. This resistance to tapering off must have be en effective during the first 2 1/2 minutes of Run 137 as well as during the later portions of the run. Therefore, Run 137 would be expected to show a greater loss of unsaturation than a run which had the same initial concentration of complex, but no excess perchloric acid. The magnitude of this effect could not be ascertained, but qualitatively it wculd be in the direction of making the .0.00095 Mor 0.00105 M value for Run 137 comparable to a value of 0.00070 M for the initial drop in Runs 129-136. It could, therefore, be concluded that the formation of one mole of the complex required one mole of perchloric acid. The discrepancy in the case of Run 138 could be much more easily resolved. The da_ta for tests Ml-M3 (Table 36) showed clearly that, in the presence of mercuric acetate, hexyne was converted to an alkene even in the absence of perchloric acid. In these runs, too, the results are somewhat random. The best value for the amount of hexyne reacted in 2 1/2 minutes in the presence of 0.0010 M mercuric acetate was 0.00035 M. Subtracting 0.00035 M from 0.00110 M gave 0.00075 Mas the amount of hexyne reacted in 2 1/2 minutes in Run 138 by way -124- Table 36. The Rate of Reaction of 3-Hexyne with Acetic Acid. Catalysis by either Mercuric A~etate or Perchloric Acid. Test No. Temp. {Deg.) HCl0i M x o3 Time Min. Hexyne M x 103 Al 27 1.5 0 2.5 58 (5.80) 5.63 5.85 A2 27 1 0 2.5 26 (5.80) 5.55 5.75 A3 25 2.5 0 2.5 32 (5.80) 5.55 5.55 Ml 23 0.5 0 2.5 32 ( 5. 80) 5.66 4.50 M2 23 1.0 0 2.5 20 (5.80) 5.26 5.35 M2b 24 1.0 0 2.5 54 (5.80} 5.49 5.31 0 2.5 26 (5.80) 5.26 5.38 M3 23 Hg(0Acl2 M x 10 1.5 -125- of the complex. This value was in close a greement with the value of 0.00070 M which was to be expected if one mole of the complex included one mole of mercuric acetate. The kinetic data, then, can be interpreted only 6n, the pasis that the reactive species is a complex including mercuric acetate and perchloric acid as well as hexyne. The fact that the rate of reaction is much greater in the presence of both catalysts than of either one alone shows that the complex must be formed relatively rapidly. Kinetic studies, moreover, furnish the information that this complex includes one mole of mercuric acetate and one mole of perchloric acid. Independent corroboration is thus furnished for the indicator studies which also demonstrated the existence of this complex , and showed th a t its formation was complete in less than one minute (Tables 31 and 32). From the indicator measurements the additional information is available that the c omplex includes one mole of hexyne. The mechanism of the catalytic addition of acetic acid to 3-hexyne, therefore, has been established. It is represented in Equations 49 and 50. The rate-determining step is sho wn in Equation 50. Et-C:C-Et .f Hg(0Ac) 2 ♦ H" Et-c;c-Et•Hg0Ac 4 .; H0Ac Et-CH:C-Et .f Hg(0Ac) 2 .f H* 1 0Ac As h a s already been pointed out, the exac• nature of the 1:1:1 complex of hexyne, mercuric acetate, and pe rchloric acid has (49) (50) -126- not been dem0nstrated. The hypothesis that the complex is a cation, however, is especially w~ll suited to this mechanism. The driving force of Equation 50 would then be the electrostatic attraction of this cation for electron-rich oxygen atoms of the acetic acid. B. Special Tests Four special runs were made to test, ~espectively, the effect of a high catalyst concentration and the effect of adding acetic anhydride, water, and diethyl ketone on the rate of addition of acetic acid to hexyne. The data are shown in Table 37. The run with high concentration of catalysts was Run 139. It is seen tha t the unsaturation dropped quickly to half its original value and then remained constant. (The erratic nature of the results may have been due, in part, to the fact that suction, as well as nitrogen pressure, was used in taking the aliquots.) This result, barring a coincidence, must represent the immediate reaction of the alkyne to g ive an alkene or alkenes. The static nature of the solution after the initial drop confirms the assumption, previously made, that the rate of the reaction of hexenyl acetate to form hexanone is slow compared to the rate of reaction of hexyne. Runs 140 through 142 resemble Run 131 with respect to initial concentrations of catalysts and hexyne. Run 131 showed a loss of 0.00150 moles per liter of hexyne in the first 2 1/2 minutes. The corresponding figures for Runs 140~142 are: ... 127... Table 37. The Effect of Special Conditions on the Rate of Addition of Acetic Acid to 3-Hexyne. Temp. Hg(OAc)2 HCl04 Additive Time Hexyne Unsaturation N x 103 Min. M x 103 (Deg. ) M x 103 M x 10 3 M Run No. 139 29 10 10 140 25 1 1 141 25 1 142 26 1 co a. Et b. H2 o c. Ac20 2 0 1.5 9 28 120 155 (3.48) 1.9a 0 2 11 114 (5.80) 4.15 3.57 3.04 1 0.277b (5.80) 0 4.51 2.5 12.25 4.11 57 4.02 1 0.265c 0 2.5 10 45 (5.80) 4.11 3.66 3.17 (13.9) 6.9 6.5 6.6 7.5 6.9 -128- 140 (with added diethyl ketone), 0.00170 moles per liter; 141 (with added water), 0.00129 moles per liter; 142 (with added acetic anhydride), 0.00169 moles per liter. All three figures probably differ just enough from 0.00150 moles per liter to represent a significant difference from Run 131. It was, therefore, tentatively concluded that diethyl ketone (in a relatively high concentration) and acetic anhydride exert an accelerating effect and water a retarding effect on the rate of addition of acetic acid to hexyne. In Run 141, with added water, the amount of yellow color formed is much less than usual, and it may be that the reason for the slower rate observed is interference with the formation of the by-product. It should be noted that the effects caused by the water and the acetic anhydride are rather small. The amount of water in Run 141 is about the same as in commercial, C.P. ac~tic acid (m.p. 15.6°), and the effect on the rate is only about 15%. It is concluded that the early kinetic experiments, in which excess ·water or acetic anhydride were often present, were not seriously effected by these contaminants. C. The Correspondence of the Amount of Yellow By-product to the Loss of Unsaturation. Runs Cl30 through Cl38, which appeared in Table 15 and were discussed in Section IV A5, were exact duplicates of Runs 130"138 just discussed. From the data of Table 15, it was learned that, in most runs, the amount of color produced -129- was proportional to the concentration of the hexyne-mercuric acetate-perchloric acid complex. Since the amount of unsaturation lost is also proportional to this concentration, it follows that, in most runs, the amount of yellow color is proportional to the loss of unsaturation. The possibility must be considered, there- fore, that in measuring the rate of change of unsaturation, it was the rate of production of the yellow species rather than of hexenyl acetate that was being determined. In Section II the preparation of 6 1/2 ml. of hexenyl acetate from 40 ml. of hexyne was reported. If correction is made for the fact that only 1/3 instead of 1/2 of the unsaturation ~as lost in the synthetic media, the indicated yield on the basis of the hexyne reacted is about 25%. Correction for material left in the water-acetic acid solution during the steam distillation and for product lost into the Dry Ice trap during the vacuum distillation would further increase the indicated yield. On the other hand, the product obtained was not pure. The possibility must be considered that the yellow material, and not hexenyl acetate, is the major product formed during the reaction. The formation of hexenyl acetate, however, even in a yield of only 25%, must have had an effect on the rate of loss of unsaturation. The data of Runs 129-138, from which it was concluded that the rate of loss of unsaturation was proportional to the concentration of the complex, covered a 3-fold change in the initial rate. The total concentration of hexyne was also changed more than 3-fold in these runs. Had the hexenyl acetate -130- been formed by some other mechanism, it . probably would not have been found that the rate of loss of unsaturation was proportional to the concentration of the complex in each of these runs. There were, moreover, two runs in which the amount of color formed was not proportional to the rate of chan g e of unsaturation but in which the rate of change of unsaturation was nearly proportional to the concentration of complex. In Run 141, with 0.277 ~ of water, the yellow-color reached less than 1/3 of the optical density of Run 131. The initial rate of change of unsaturation, however, a greed within 15% for the two runs. Of greater significance is a comparison of Run Cl38 and 138. These runs were 0.0015 Min mercuric acetate and 0.0005 Min perchloric acid. It will be recalled that no color at all was formed in Run Cl38, but that the initial rate of loss of unsaturation in Run 138 was greater than in Run 133b. After correction was made for the catalytic effect of the excess mercuric acetate in Run 138, the initial chan g e of unsaturation of Runs 133 and 138 a greed closely despite the absence of yellow color in Run 138. It is, therefore, concluded tha t the yellow color, if its formation coincides with a chan g e in unsaturation, is formed only in very small quantity. It is further concluded that the mechanism elucidated for the rate of ch ange of unsaturation may be safely ascribed to the formation of hexenyl acetate. -131- D. The Rate of the Reaction of the Hexyne-Mercuric Acetate-Perchloric Acid Complex with Acetic Acid. It was shown to be possible to arrange Runs 129-138 into three groups on the basis of the initial rates of reaction and the concentration of the 1:1:1 co mp lex in these solutions. The first group (0.0015 moles per liter of complex) included Run 130; the second group (0.0010 moles per liter of complex) included Runs 129, 131, 132, and 135; the third group (0.0005 moles per liter of complex) included Runs 133 and 136 as well as Run 138 (after correction for the catalytic effect of the excess mercuric acetate). Run 137, with 0.0015 M perchloric acid and 0.0005 M mercuric acetate did not fit into these groups, havin g an initial rate somewhat greater than group three. It will be noted that the values given for the concentrations of the complex equal the concentrations of the limiting conThis approximation follows fr-om the equilibrium constant for the complex (3.5 x 107 liters2 moles- 2 ) stituent of the complex. and the relatively high concentrations of solutes in Runs 129-138. Run 136 had the lowest concentrations of solutes, being 0.0058 Min hexyne and 0.0005 Min both mercuric acetate and perchloric acid. The complex in this solution had a concentra- tion of about 0.00044 Mand thus included about 90% of each of the two catalysts. The equivalent f~action in the other runs was even higher. Runs 130, 131, 133c, and 137 can be chosen as representative of the set of determinations. A plot of the concentration of hexyne against time for each of these four runs is shown in -132- Figure 4. The difficulty of determining the tangent to these curves at zero time is immediately apparent. The "tapering-off" phenomenon is so marked that the shape of the curves at any time before the first reading is almost unknown. It is for this . reason that comparison among the runs has been made on the basis of the loss of hexyne after 2 1/2 minutes. A. further disconcerting fact in connection with determining the true initial rates is the lack of certainty that the complex titrates as an alkyne. It is conceivable that only two equiva- lents of bromine add to each mole of complex. Even if this were true, the g rouping of Runs 129wl38 and the proof of the mechanism of the reaction would not be cha nged. The initial, 2 1/2 minute loss of unsaturation would be 0.0075 moles per liter, 0.0050 moles per liter, and 0.0020 moles per liter respectively for the three groups - values which are still proportional to the conce ntrations of complex in each case. The apparent initi~l rates, howe ver, would be considerably decreased. On the grounds that rcugh approximations may be better than no values whatever, two estimates will be made as to the value of kin Equation 30. For the first value it will be - dH • k CH dt (30) assumed, as seems most likely on chemical g rounds, that the complex titrates as a triple bond. The approximation will then be made tha t the initial slope is twice as great as the slope that would be obtained by drawing a straight line between the initial concentration and the 2 1/2 minute point. The three 0.002 00 0 . 00300 0 . 00400 0.00500 0.00600 moles/liter Hexyne Cone . of 0 10 20 --------30 40 -- 50 Tim e .i n mi n utes 60. Run 130 · Run 137 Run 131 Run 133c The Cha n ge of Concentration of 3-Hexyne in Acetic Acid Solutions containing Mercuric Acetate and Perchloric Acid. Figure · 4. I 01 01 I ..... -134- values of k for the three groups of runs will then be: (0.00225)(2)/(2.5)(.0015); (2)(.0005). case is k (0.00155)(2)/(2.5)(.001); (0.00075) The value calculated for the rate constant in each . -1 = 1 min. On the other hand, the hypothesis m~y be made th&t the complex titrates as an alkene. In this case the abrupt tapering-off phenomenon would be explained away. There would be simply an immediate drop in unsaturation caused by the formation of the complex followed by a nearly. linear succession of points on the unsaturation vs. time gr~ph. The initial rate would be about l/6th of the value calculated above, or, approximately, k E. = 0.2 min.- 1 The Rate of Reaction of 3-Acetoxyhexene-3 with Acetic Acid. Catalysis by Mercuric Acetate and Perchloric Ac id. · As a final step in the study of the rate and mechanism of the addition of acetic acid to 3-hexyne, a few runs were made to study the rate of addition of acetic acid to hexenyl acetate. The runs were made under approximately the same con- ditions as Runs 129-138, their purpose being to confirm the hypothesis, drawn from Run 139, that the reaction rate of hexenyl acetate is slow compared to the reaction rate of hexyne. Pre- cautions to exclude oxygen were taken as in Runs 129-142. The data are shown in Table 38 and Figure 5. It is evident that, Run 139 notwithstanding, the reaction of hexenyl acetate with acetic acid proceeds ~ta rate which is approximately equal to the rate of reaction of hexyne. Runs 132 and 205 -135- have the same concentrations of catalysts and ab out equal concentrations of hexyne and hexenyl acetate respectively. In Run 132 0.0014 moles per liter of hexyne (or else 0.0005 moles per liter - depending on the hypothesis take n for the effective unsaturation of the complex) reacted in the first two and a half minutes. In Run 205, the loss of hexenyl acetate in the first 2 1/2 minutes was 0.0008 moles per liter. Nevertheless, it is probable the.t the initial rates o bserved in Runs 129-138 were not significantly affected by the reaction of hexenyl acetate. In Runs 203-207, the abs-e nce of p_exyne left the catalysts free to catalyze t he reacti on of the vinyl ester, whereas in Runs 129-138 the relatively large excess or· hexyne would have effectively complexed the catalysts and inhibited the reaction of hexenyl acetate. This set of data may be examined for the information it gives on the mechanism of the catalytic addition of acetic icid to 3-acetoxyhexene-3. It is evident, to begin with, that these runs are nearly free of the tapering-off phenomenon present in similar determinations with hexyne. The runs do taper-off slightly - both from zero-order dependence on the concentration of the ester, as shown in Figure 5, and from first-order dependence, as can be seen if the data is plotted on semilogarithm paper. It is probable th at the products (in this case ethylpropyl ketone and acetic anhydride) exert an inhibiting effect. Comparison of Runs 203 and 205 suggests that the initial rate of reacti on depends on the concentration of the -136- Table 38. The Rate of Re a ct i on of He xenyl Acet a te and Acetic Acid. Catalysis by Mercuric Acetate and Pe rchloric Acid. Time Min. Hex enyl Acetate M x 103 0.56 0 2.5 10 39 110 (25. 0) 24.8 24.9 24.4 24.8 1 1 0 2.5 11 42 76 180 (27.0) 26.0 23.8 17.3 12.1 6.l 0.28 25 l 0.4 0 2.75 11.8 49 105 191 (27.0) 26.4 25.8 . 23;·3 20.8 18. 4 0.10 26 1 1 0 2.3 10.5 33 49.5 (13.5) 0.22 Run No. Temp. (Deg. ) Hg(OAci2 M x 10 202 25 00 203 25 204 205 HClO M x fo3 Initial Slope x 103 12.8 11.1 7.0 5.2 206 25 0.4 1 0 3 10.5 37.5 63 (27.0) 26.7 25.0 20.4 17.1 0.18 207 26 0.4 0.4 0 2.25 42.2 76 116 (27.0) 26.4 25.0 22.9 21.8 0.045 208 27 1 0 2.5 9.5 (27.0) 26.3 25.9 25 . 9 41 ·- -137- hexenyl acetate only very slightly - somewhere between zero and first-order dependence but closer to zero-order. This result would be expected if the reaction took place b y way of a complex the equilibrium constant of which was moderately large. Increas- in g the concentrati on of ester would then cause some increase in the concentration of the complex. The data of Runs 203, 204, 206, and 207 indicate that the rate of reaction was dependent on the concentration of both the perchloric acid and the mercuric acetate. The dependence on the concentration of perchloric acid was somewhat greater than first power and on the mercuric acetate s om ewhat less than first power. In this case the absence of a complex or else a complex with a very low equilibrium constant seemed to be indicated. Increasing the concentration of either the mercuric acetate or the perchloric acid caused an increase in the observed rate, whether or not the two catalysts were pres e nt in equal concentrations. Because of these divergent indications, these runs are not considered sufficient to establish the mechanism. The indicator studies of the hexenyl acetate-mercuric acetate complex also led to results which were not found to be interpretable. It seems evident that there must be a relationshio • between the two se ts of data, but it has not been found. 0 0 C 0 ~ .µ ~ st) +:> •.-4 0 C 0 lo-! ti) I -a: 0 (I) .µ 0 >< ~ ..c: >< «) (I) C G) I ti) 0 0 . 0 050 0. 0100 0 . 0150 0.0200 0.0250 0.0300 0 20 40 Run 205 60 ~~ '---, Run 206 80 Fi g ure 5 Run 203 100 Tim e in min ut e s The Ch e ~g e of Concentration of 3-Acet o xyhexe ne- 3 in Acetic Acid Solutions c ontaining Merc uric Acetate and Per c h lo ric Ac id. - Run 204 Run 207 I Ill ..;J . I r' -~ -138- VII. Summary The reaction of 3-hexyne with glacial acetic acid has been studied in acetic acid. In the presence of a mercuric salt and a strong acid, two moles of acetic acid were found to add to the triple bond of the hydrocarbon, fcr~ing, successively, 3-acetoxyhexene-3 (hexenyl acetate) and 3,3-diacetoxyhexane (3-hexanone diacetate). The isolation of 3-acetoxyhexene-3 and 3-hexanone was accomplished by steam distillation of particular solutions. Because the hexanone might have been produced by hydrolysis of the hexenyl acetate during isolation, two subsidiary proofs were offered that the second mole of acetic acid doe,s add. The acetic acid solution from which the ketone was isolated lost 95% of its initial unsaturation before the steam distillation was begun. Moreover, appropriately catalyzed solutions of hexenyl acetate in acetic acid were themselves found to lose unsaturation. The isolation of the hexanone rather than its diester is due to the instability of the diester, not only during the steam distillation but also, probably, in the acetic acid solution. · The reaction solutions were usually catalyzed both by mercuric acetate and by one of three acids - perchloric acid, sulfuric acid, or boron trifluoride• was followed by bromine titration. The course of each reaction Inconclusive results were secured using boron trifluoride, perhaps because this acid is an unusually strong electron acceptor and, therefore, more -139- affected by impurities than the other acids. Many data w:er-e collected from runs in which sulfuric acid was the acid catalyst. No mechanism was ever found to interpret the:sedata. By means of indicator studies it was learned that the achdity of acetic acid solutiops of sulfuric acid is proportional to the concentration of the s ulfuric acid at dilutions g reater than 0.2 M, but that the acidity of these solutions is much less than the acidity of solutions containing equal concentrations of perchloric' acid. Since no explanati on for this anomaly was known, the studies using sulfuric acid were abandoned. The mechanism of the reaction was finally elucidated from a set of runs using perchloric acid as the acid catalyst. It was found that a complex of hexyne with one mole of mercuric acetate and one mole of perchloric acid was rapidly formed in the solutions and th a t this complex (probably a hexyne-mercuric monoacetate cation) was the rea c tive species. Catalysis by either mercuric acetate or perchloric acid in excess of the complex did occur but at a slower rate. The rate of the addition of acetic acid to this complex is in doubt be c ause of the uncertainty as to whether the complex titrates as an alkyne or an alkene. The first order rate con- stant is probably between 0.2 a nd 1 min- 1 • Some data were collected on the rate of the reaction of hexenyl acetate with acetic acid in the presence of mercuric acetate and perchloric acid. Smooth, almost linear, curves of -140- concentration against time were obtained, but no satisfactory interpretation as to mechanism was · deduced. Indicator studies were undertaken in an attempt to secure information concerning the hexyne-mercuric acetate complex. It was found that the complex included one mole of hexyne in addition to one mole each of mercuric acetate and perchloric acid. The equilibrium constant was found to be 3.5 x 10 7 moles- 2 liters 2 • A similar s e t of investigations with hexenyl acetate proved t h at complexes were f or med with this ester but no equilibrium constant could be fo und which would fit all the data. The indicator studies were extended to the investi g ation of weak bases. The following activity constants (pKa) were calcu.... lated for seven weak bases fr om observations made in acetic acid: -- acetonitrile, -4.2; dioxane, -4; p-toluenesulfonamide, w3.2; water, -2.35; acetanilide, -0.9; acetamide, 0.1; urea, 0.5. -141- References 1. Hennion, G. F., Vogt, R. R., and Nieuwland, J. A., J. Org. Che~ . , ! , 159 (1936). 2. Wright, G. F., J. Am. Chem. Soc., 57, 1993 (1935). 3. Frieman, R. H. Kennedy, E. R., and Lucas, H. J., J. Am. Chem. Soc., 59, 722 (1937). - 4. Sand, J., and Breest, F., 5. Hugel, G., and Hibou, J., Chimie et industrie, Special No. 296, Feb. 1929. C.A. 23-,-3898 (1929). 6. Adams, R., Roman, F. L., and Sperry, W. N., J. A.m. Chem. · Soc., 44, 1781 (1922) and footnotes. See aiso Ref. ?and footnotes. 7. Lucas, H. J., Hepner, F. R. and Winstein, S., J. Am. Chem. Soc., 61, 3102 (1939). 8. Hennion, G. F., and Nieuwland, J. A., J. Am. Chem. Soc., 56, 1802 (1934). 9. Bried, E. A., and Hennion, G. F., J. Am. Chem. Soc., 59, 1310 (1937). 10. Vaugn, T. H., Hennion, G. F., Vogt, R. R. and Nieuwland, J. A., J, Org. Chem., 2, 1 (1937). 11. Lespieau and Wiemann, Bull.Soc. chim., (4) 45, 627 (1929). z. physik. Chem, 59, 424 (1907). Huntress, E. H., and Mulliken, R. S., Indentification of Pure Organic Compounds, New York, 1941. 12. Campbell, K. N., and Eby, L. T., J. Am. Chem. Soc., 63, 2683 (1941). 13. Henne, A. L. and Greenlee, K. W., J. Am. Chem. Soc., 67, 484 (1948). 14. International Critical Tables, Vol. IV., 1st Ed., New York, 1928., p. 183. - - -- 15. Todd, F., Ind. Eng. Chem., Anal. Ed., 17, 185 (1945). 16. Eberz, W. F., and Lucas, H. J., J. Am. Chem. Soc., 56, 1230 (1934). -142- References (Continued) 17. Kolthoff, I. 1\11., J. Am. Chem. Soc., 48, 1447 (1926}. 18. Handboo~ of Chemistry and Physics 29th Ed., Cleveland, 1945., p. 1553. 19. LllJnge Handbook of Chemistry IVth Ed., Sandusky, Ohio, 1941., p. 1580. 20. Glasstone, S., Textbook of Physical Chemistry 2nd Ed., New York, 1946., pp-.-529-530. 21. Sabatier, P. and Mailhe, A., Compt. rend., 156, 1730 (1913). 22. Eisenlohr, F., Zeit. physik. Chem., 75, 585 (1910). · 23. Brandstatter, M., Mikrochemie ver. Mikrochim. Acta, 32, 33 (1944). C.A. 40,262 (1946). -- 24. Allen, C. F. H., J. Am. Chem. Soc., 52, 2955 (1930). 25. Whitmore, F. C. and Karnatz, F. A., J. Am. Chem. Soc., 60, 2533 (1938). 26. Frieman, R.H. Kennedy, E. R. and Lucas, H. J., J. Chem. Soc., 59, 722 (1937). 27. Lucas, H.J., and Pressman, D., Ind. and Eng. Chem., Anal. Ed. 10, 140 (1938). - 28. Swift, E. H., A System of Chemical Analysis, NewYork, 1946., p. 82. Arn. 29a. Hofmann, K. A. Zedtwitz, G. A., and Wagner, H., Ber. 42, 4390 (1909). b. Hofmann, K. A., and Kirmreuther, Ber. 42, 4856 (1909). 30. Flexser, L. A., and Hammett, L. P., J. Am. Chem. Soc., 60, 885 (1938) and previous papers by Hammettand cbllaborators. 31. Hall, N. F., and Spengeman, W. F., J. Am. Chem. Soc., 62, 2487 (1940). - -- - 32. Hammett, L. P., Physical Organic Chemistry, NewYork, 1940. -143- References (Continued) 33a. Hammett, L. p. ' and Deyrup, A. J.' J. Am. Chem. Soc., 54, 2721 (1932). b. Hammett, L. p. ' and Deyrup, A. J.' J. Am. Chem. Soc., 54, 4239 (1932). 34. Holleman, A. E.' and van Haeften, F. E. ' Rec. trav. chim. , 40, 67 (1921). 35 . Behrens, H., Rec. trav. chim., 21, 269 (1902). 36. Ting le, J. B., and Blanck, F. C., J. Am. Chem. Soc., 30, 1395 (1908). - -- 37. Merz, V., and Weith, W., Ber., 10, 746 (1877). 38. Drake, N. L., a nd 10 collaborators, J. Am. Chem. Soc., 68, 1602 (1946). - 39. Blanksma, J. J., and Schreinemachers, H. H., Rec. trav. chim., 52, 428 (1933). -- --- 40. Gillis, J., and Delaunois, A., Rec. trav. chi m. , 53, 186 (1934). 41. Me rckx, R., Verhulst, J., and Bruylants, P., Bull. soc. chim. Belg., 42, 177 (1933). -- -- 42. Hammett ·, L. P., and Paul, M. A., 827 (1934). 43. See 44. Hal 1, N. F., J. Am. Che m. Soc ., 52, 5115 (1930). i• Arn. Chem. Soc., 56, references in Taylor, T. W. J., J. Chem. Soc., (1930), 2741. ....144_ Propositions I. Bartlett and Dauben f ound that trichlorophenol was a stron ger acid but a weaker h yd roge n donor than other phenols which they investigated. an omaly is probably wron g . Their explanation for this apparent A si mpler ex planation is proposed, based on . the tendency of the ortho-chloro atoms to prevent either the approach of the h ydro gen ion to the phenolate oxygen atom or the approach of a base to the phenol group. Bartlett, P. D. and Dauben, H. J., J. Am. Chem. Soc., 62, 1339 (1940). II. The effect of the presence of mercuric salts on the ozonolysis of alkenes and alkynes should be investigated with a view to increasing the yi e lds and the rates of the reactions. III. It is proposed that the reactive intermediate for aldol con d ensations occurring in water is an anion of the form sho wn here. Remick, A. E. Electronic Interoretations of Organic Chemistry, New York , 1949., p.~466. IV. a. The steric inhibition of resonance is an accepted explanati on for the enhanced basicit y of N-alkylated, -145- Propositions (continued) ortho-substituted anilines over the free, ortho-s~bstituted anilines. It is proposed that the same effect also plays a part in the greater basici ty of N-alkylated anilines as compared to aniline itself. b. The compound produced by the alkylation of an aromatic amine is usually a stron ge r base than the parent amine. An exception is 4-chloro-2-nitro-N-methylaniline. An explana- tion, based on the B-strain hypothesis, is proposed for this phenomenon. H. C. Brown, 67, 1452 (1945). v. a. Acidimetric measurements in acetic acid provide a promisin g tool for studying the complexes formed by metallic ions and organic compounds in solution. The investigation by this method is proposed for saturated carbocyclic and unsaturated aliphatic hydrocarbons, for aromatic hydrocarbons, including carcino gen ic materials (proposed in conference with Prof. Lucas), for carbonyl compounds, and for esters. b. The electronic con.fi g urations of the atoms which most readily for m metal-olefin coordination complexe s suggest that a full d subshell underlying the valence electrons may be a characteristic feature of the me t allic atom in these complexes. It is proposed that barium, one of the rare earch me tals, and gold be investi g ated for their ability to form such complexes. Keller., H. N., Che m. Rev., 28, 229 (1941). -146- Propositions (continued) VI. The eff ect of t he pr e sence of me rcuric salts on the rate of bromination of ketones should be investigated. Thesis, pp. 37-38. VII. The indicator method of measuring the basic strength of weak bases in acetic acid should be a useful tool for the investi gation of the following problems: 1) the structure of 2 certain aliphatic amides 1 ; 2) the dibasic nature of acetamide ; 3) the rate of the acid-catalyzed esterification of alcohols; 4) the effect of ring-strain on the basic strength of ethylene- and polymethylene- oxides; 5) the effect of a positive pole on the basic nature of an amine gr oup in the same rriolecule (idea for this problem suggested by Prof. Niemann); 6) the basicit y of pyrrole, isocyanates, and mesitylene 3 • 2. 3. Cason, J. and Wolfhagen, H. J., J. Org. Chem., 14, 155 (1949). TayI'or, T. W. J., J. Chem. Soc., (1930) 2741. Copley, M• . J., Zellholfer, 0:-:-F., and Marvel, C. S., J. Am. Chem. Soc., 61, 3550 (1939). a. An investigation of the ability of butadiene 1. VIII. to catalyze the conversion of parahydrogen to orthohydrogen would permit evaluation of the postulate 1 that one of the resonance forms of butadiene has two unpaired electrons. 2 1. 2. Paulin g , L., The Nature of the Chemical Bond, Ithaca, l94~p. 217. - - Selwood, P. W., Magnetochemistry, Evanston, 1943. -147- Propositions (continued) VIII. A surprising ,(, f b. rearrangement has been reported. 1 fl -complex 2 -migration in a pinacalone It is proposed that a is formed between the carbon-carbon triple bond and the migrating phenyl group. 1. 2. IX. Kleinfeller, and Eckert, Ber., 62, 1598 (1929). Dewar, M. J~ S., The Electronic Theory of Organic Chemistry, Oxford, 1949. The syntheses of the hydrocarbons and would be accomplished b y the decarboxylation of the corresponding 1 tricarboxylic acids . It is proposed that this decarboxylation be attemped by treatment with silver oxide and brom~ne 2 • 1. 2. X. Thorpe, J. F. and Beesly, R. M., J. Chem. Soc., 117, 591 (1920). Hunsdiecker, H., and Hunsdiecker, Cl., Ber. 75, 291 (1942). On the basis o f currently available inforrration, the production cf a hydro g en bomb seems highly improbable. Time, Dec. 12, 1949., p. 48.